searchlore

0:00

[Music] Hey, hey, hey.

0:25

[Music] [Music] Hey, hey, hey. [Music] [Music] Heat. Heat.

2:08

[Music] [Applause] [Music] [Music] Heat. [Music] Heat. Heat. Heat.

2:42

[Music] [Applause] [Music] We came to this [Music] world to reach the stars.

3:14

We came to this [Music] world to see a future.

3:26

You came to this world to reach the stars to shape our future.

3:41

We came to feel the [Music] new Heat. Hey, heat. Hey, heat.

3:58

[Music] [Applause] [Music] [Music] [Applause] [Music] Let [Music] me feel.

4:39

[Music] [Applause] [Music] You're watching TVPN.

4:50

Today is Tuesday, May 20th, 2025.

4:52

We are live from the Temple of Technology, the Fortune Finance, the capital of capital, and we are recapping Microsoft Build, which was yesterday, and Google IO, which is today.

5:01

Uh we have a bunch of folks coming on from Google.

5:02

Uh and but we're going through what happened at Microsoft.

5:07

Obviously, they're going even deeper into artificial intelligence.

5:09

uh talking about the agentic web, the future where the robots do your work for you and uh work while you sleep.

5:16

Um Google, this is why we do three hours a day, folks. Yep. So we can cover it all.

5:23

We're going to have time to watch to the full thing every single day.

5:26

Crazy amount of news and deep dives this week specifically into AI.

5:31

Obviously Microsoft and Google are doing big keynotes right now, but uh Bloomberg has a fantastic deep dive into OpenAI's Stargate.

5:41

And then there's also a profile by the same journalist in Bloomberg about uh Daario O Amade over at Anthropic telling his soul story.

5:48

And so uh basically everyone's getting coverage uh and AI is the only thing people care about.

5:56

And so that's what we're talking about today on the show.

6:01

Um, but let's give you a little bit of a recap of the news.

6:03

So, Microsoft Build 2025 opens.

6:06

Satcha Nadella calling for an open agentic web.

6:08

Uh, there's new co-pilot features, open protocols.

6:12

They're going all in on MCP uh model context protocol that uh standard not a particular tech product obviously defined by uh anthropic but then open sourced and then adopted by both Microsoft and Google.

6:24

So, uh the narrative is MCP1 basically.

6:26

uh and MCP became a very hot button issue.

6:30

Lots of people posting threads on how to use it.

6:33

It's goated yet, but it's definitely in the conversation now.

6:35

It seems it seems like it could be goated. It's doing quite well.

6:38

It's getting massive adoption.

6:39

And most importantly, I think is like it's not it's more of a standard than a particular product.

6:43

And so you're not seeing a wave of startups, but you're seeing startups use MCP.

6:46

Uh and then Elon Musk joined Sachi Nadella uh on stage via Zoom uh presumably or or or probably Microsoft Teams, I guess.

6:55

they called in from um and confirmed that Azure will host XAI's Gro 3 and Gro 3 Mini.

7:03

Uh and he kind of talked about his vision for um Grock being a truth seeeking LLM and focused on foundation uh a foundation model built on physics and first principles which has been his uh his goal with that.

7:18

But if you look at Sachin Nadella's uh pitch now for the host the models that are hosted on Azure, it's really just everything you can run.

7:27

They even have so they have a you can swap quickly between even DeepSeek is on there and there's a version of Deepseek that's been fine-tuned by Microsoft's AI internal team.

7:38

There's all the llama models.

7:40

There's all the open AI models.

7:40

Now the Gro models it out.

7:42

Yeah, maybe um they found him. They asked him to leave. Now they're hosting it. Good. Yeah.

7:48

Um and and then there's even there's so so many models now that that Azure has a model router for OpenAI models.

7:56

So you can import the Azure OpenAI model router and and then it will decide what the cost benefit is for the individual model that you're using.

8:08

Oh, do you have a do you have a plug or I guess we I guess we Oh, no. I got this one. Cool. Okay.

8:14

Um and so yeah, you should go watch the full uh conversation between uh Microsoft CEO Sachi Nadella and Elon Musk. He says with Grock 3.

8:23

5, which is about to be released, it's trying to reason from first principles.

8:26

Uh also I saw a ton of uh Tesla battery pack show.

8:30

Bad news for product managers who have been Oh yeah.

8:34

You know, spending their entire life figuring out how to think from first principles that that rock 5 will just do it out of the box. Yeah. Yeah.

8:39

It's a golden age for people that never learned first principles.

8:42

You can just say think from first principles. Yeah, do it for me. Yeah, just do it.

8:46

Golden retriever mode all over again. Bull market.

8:47

We've been calling it bold prediction at the time. Dumb.

8:51

Uh Dan Shipper summed it up.

8:55

Yesterday Microsoft rolls out MCP to support support to Windows.

8:57

Today Google officially supports MCP and the Gemini SDK. It's over. MCP wins.

9:02

Um and that makes a lot of sense.

9:04

Uh it's a standard that everyone's building against now.

9:07

And uh we we imagine that that will continue because once you have Google and Microsoft and Anthropic, everyone has to conform to it.

9:13

Coordinating with Dan, but hoping to get him on the show Thursday. He's got some news.

9:17

Um another big announcement from Google IO.

9:21

They have real time Google Meet translation is live now.

9:27

I feel like we've seen this demo 10 years in a row.

9:29

Um but it feels like it's finally here.

9:33

Uh, and so there's a uh a video where you can be talking to somebody in a different language and it will live translate what you're saying into the local language.

9:43

So you can just have a conversation.

9:45

Obviously, this is a it's like one of the most obvious AI use cases.

9:49

It's something that Google should do.

9:51

It's something that seems like extremely beneficial for for humanity across the board. Just like pure alpha. So excited to see that.

9:59

They're also releasing a diffusion language model that I want to dig in with the Google team more on because um we've seen image models move away from different into more token based structures and transformer based architectures.

10:13

Uh now language models are going into fusion.

10:17

So I don't know enough about this but I'm going to be digging in this week.

10:19

Uh they also launched deep think 2.

10:20

5 pro uh new enhanced reasoning mode for using uh Google's research in parallel thinking techniques meaning it explores multiple hypotheses before responding um basically branching all the different trees.

10:35

There's already a little bit of this going on in OpenAI's deep research product but obviously it's a very competitive space.

10:40

Uh this enables it to handle incredibly complex math and coding problems more effectively.

10:45

Um, and I was excited about Google Gemini.

10:53

I've become more bullish on the product since I sent you those incredible AI generated videos that are generated from V2.

10:59

And so I went in to Gemini because remember I I generated three or four of those videos and then it gave me a timeout and it said, "Look, I know that you've spent millions of dollars with Google over your career, course of your career, but $20 is the most money I could accept from you.

11:15

I couldn't possibly charge you an extra dollar to make one more of these videos.

11:19

200 just doesn't do the GPUs are on fire.

11:20

Even though you would pay 2,000 and we have even though we have a hundred billion dollars of cash, we have to rate limit you.

11:29

Yeah, we couldn't possibly light another GPU on fire for you right now. There's no price.

11:33

We should ask Logan later around rate limiting and and the plans there.

11:36

Hopefully it was a very frustrating experience for me because I would have paid more in that moment to generate one more photo of a Lamborghini Urus with a TVPN livery driving around a track.

11:46

Uh but I couldn't do it and and it was it was not possible.

11:48

I mean I guess I probably could have if I got a separate Google account with a separate Gemini Pro subscription maybe on a separate phone and then had to do that. Yeah.

11:58

huge opportunity right now to create a a basically a bundled Google subscription where you subscribe to a service that then aggregates like a,000 you know it's kind of like how there's how the banks uh have FDIC insurance and you're insured for up to like $100,000 then $250,000.

12:15

So some people would take you know oh spread spread a million dollars across 10 accounts or five four accounts.

12:22

You need to do that with Google so you don't hit the rate limits clearly.

12:24

Um so they put me on a timeout.

12:27

They said, "John, you you have generated too many of these uh of these GT3 RS's, TBPN liveries.

12:32

Um, you need to take a break.

12:35

We'll see you one day from now." Okay, one day. You can come back.

12:37

It's been three days now since I did that.

12:42

That You should be okay with that because that's kind of a typical workflow with a human, right?

12:45

You maybe get a few iterations and they say, "Hey, I got to go to sleep. I got to take a break. I'll see you tomorrow." Yeah.

12:51

You're overworking me, boss. Yeah.

12:51

And so, it told me to take a break.

12:53

It said, "One day we'll see."

12:55

the agentic web Monday night. I remember what it said.

12:58

It said Monday night you can create more.

13:00

So I wanted to go and create one today because it's Tuesday and I said generate a vertical video 16 by9 of Cinder Pai arriving at Google IO in a Whimo and waving to a crowd of adoring fans who are taking pictures with Android phones.

13:10

And it says too many requests in a short time period. Try again later.

13:14

So again a weird thing where like the product is amazing but I can't access it enough and I keep being frustrated with the actual roll out of these products but hopefully it's getting better.

13:25

I do wonder if they're getting slammed right now because of Google IO.

13:28

Everyone's testing out the new model and so they really are under uh a ton of uh demand.

13:31

What was interesting is that it it didn't generate any videos for me today and immediately it just said it just said too many requests in a short period of time.

13:42

It didn't say if that if that's too many requests from me because this is the first request I've sent in today and so I was very unsure of what's going on there.

13:50

Is it just their systems under load?

13:51

Can you prompt it to address you as boss man?

13:53

because it might be more it would feel a little bit better if it said, "Hey, boss man, that's enough for today."

14:00

Instead of just, you know, hitting feeling like you're hitting those rate limits. I'm working hard. Uh, what else?

14:07

Going back a little bit, one one thing that stood out to me seeing Elon and Satia having a conversation working on this product launch felt significant to me because it was not too long ago that Microsoft was named as a defendant in the OpenAI lawsuit, right?

14:24

Elon's lawsuit against OpenAI, Microsoft was a defendant.

14:27

Um, and at a certain point, I believe it was in 2023 as well, uh, Elon was threatening to sue Microsoft for training on X or Twitter at the time. I didn't know that.

14:40

And so these two haven't exactly had the best relationship.

14:45

I mean clearly like they they're fine on an individual level, but the the the partnership now uh is cool to see just given that there had been uh quite a bit of friction over the last couple years.

14:57

But the funny thing is Musk was actually way back in the day an intern for Microsoft Windows in the 90s.

15:06

So, uh, very isn't there some story about him trying to get a job at Netscape with Mark Andre and he didn't make it past like the first round and so he went to Microsoft.

15:15

That might be wrong, but there's something about uh Elon before he started Zip 2 was kind of moving around the valley at different tech companies.

15:23

It's funny to imagine intern Elon, let me look it up, writing some code for for Microsoft Office at the time or something.

15:30

Uh but now now they're partnering and I mean it's all part of Satcha's strategy to be model agnostic.

15:36

Uh it's always felt like Microsoft research has been behind the curve on LLM training like Zuck.

15:46

Although although Llama has kind of hit a rough patch with Llama 4 apparently.

15:51

So Elon uh was rejected from a job Yeah. at Netscape. Yeah, that's right.

15:55

He sent in his resume, showed up at the office.

16:01

even lingered in the lobby, but no one acknowledged him. Wow.

16:03

Just Mark Andre walking by mogged and now and now Mark's in XAI. Yeah, now they're boys.

16:11

They ripped a big check and a bunch of other companies too. But um it's great.

16:14

Um what what a full circle story.

16:18

But yeah, so I mean Satcha obviously is is positioning Microsoft to be model agnostic and wants to be a vendor of this in the cloud platform.

16:25

And I think that's it seems very good for for Azure in the idea that if you're a business and you're building on on Google, you might feel like, oh, they're going to really push the Gemini models on me.

16:38

Maybe I get locked in at some point.

16:40

Maybe that gets expensive.

16:42

But if I'm on if I'm on uh Azure, I can dance around between different OpenAI models and different llama models and different gro models and uh and and have a lot more flexibility because uh Satcha has kind of said like, hey, we're we're we're not, you know, going to push you towards a particular model or particular like regime.

17:02

They also release an open source VS code, you know, sort of like co-pilot product which is big.

17:09

Uh so anyways reaction to that was was project padawan was positive as well.

17:16

GitHub copilot graduates to a full coding agent uh project padawan uh gi for uh uh for copilot enterprise and pro plus capable for autonomous feature work bug fixes and refactors and so everyone's kind of moving into these autonomous coding agents.

17:32

We're having Scott Woo from Cognition on the show on Wednesday, tomorrow, uh to talk about the landscape and uh we're also having Lee Marie from uh Kleiner Perkins on Friday to talk about the same thing.

17:42

This idea that how is the AI coding landscape shaping up?

17:47

Is it one market, two, three, four?

17:47

It feels like there's uh as as we were talking um uh about it a little bit earlier, um it felt like there are there's AI code that's written.

18:00

I mean you just gave an example where you asked it to count you asked Chacht to count the number of this or something. What happened there?

18:08

So basically XAI got received 168 Tesla mega packs which is going to power Colossus 2 which is their second data center and uh there's this sort of above ground image and I initially saw the image and I didn't uh somebody else had had sort of counted them up.

18:27

So I asked Chat GPT to count and it and it spent 13 minutes attacking the problem by by writing code.

18:33

In that time of course I just like you know looked at uh the x-axis and the y- axis and multiplied uh and got the right number.

18:44

Um but uh chat GPT was running in the background you know attacking it in the most sophisticated way possible.

18:52

Uh ultimately what the number I wanted to figure out that was interesting is that uh the retail value of the mega packs, which of course you can buy on the Tesla website, $5 million, uh comes out to around $850 million. Wow.

19:04

Just on mega packs, which is about half of their 2024 revenue, right?

19:10

Wait, half not XI's, but X's revenue, which is now combined. Yeah. Yeah. Yeah.

19:16

XI doesn't have any X's graduated. Yeah.

19:18

But they raised a lot of money, so they're they're spending on Tesla mega packs. That's wild. Yeah.

19:24

And I'm interested to see how ultimately exactly how this new round shakes out.

19:28

There's been rumors around the new fund raise for a while.

19:33

Some final details of that round haven't been announced yet, right? It's still just rumors. Um, that's interesting.

19:36

I wonder 13 minutes is fast, but not that fast.

19:41

I wonder if there's a world where OpenAI has a function it can call to like Mechanical Turk or Scale AI to just have put a human on the case. Well, it failed. It failed.

19:51

So, I mean, it didn't even do it in 13 minutes.

19:53

It just It just basically timed out because I could see it working and writing, you know, taking all these different cracks at it. Yeah.

20:01

So, imagine to its credit, it would have taken a human, not 13 minutes to count those.

20:06

No, just manually counting them like what I did.

20:08

Not I didn't actually go line by line, but like uh you took me whatever 20 seconds.

20:17

But if uh if a human had to run take that many attacks at solving an they wouldn't like what I'm saying is that is that you kick off this how many white s how many white squares are in this image and it just displays that to a mechanical Turk who's just sitting there doing random tasks all day long and they just one-shot it in you know two minutes.

20:37

Yeah, this is what I was talking about over the weekend. I was joking around.

20:40

I'm sure some people took it seriously, but I was saying uh I met somebody in the Philippines whose job is to make investment decisions for a firm that claims to use AI.

20:48

That's the final the final boss of of of using AI. Yeah.

20:55

Uh but but it's really, you know, uh uh mechanical uh Turk is just, you know, using using uh outsourced labor to just make all investment decision.

21:05

the the overall announcement at any of these keynotes is always we're 90% of the way there and the last 10% will be another 90% of the way there.

21:15

Um but I mean there are some there are some impressive uh you know uh announcements and and steps forward but obviously there's a lot there's a lot left to do.

21:25

So Copilot Studio just running through what else uh Microsoft announced.

21:28

Copilot Studio now supports multi- aent system so developers can build agents that delegate tasks to one another with my Microsoft 365 agent builder.

21:35

Uh NL web debuts as HTML for the agentic web letting any size any site expose a conversational endpoint discoverable by AI agents.

21:44

So an AI agent shows up and it can just chat with your website.

21:48

Um, we talked about project padawan, github copilot graduating to compete in that space with codeex from openai and now just in the openi world you have three products that effectively write code.

22:02

You can just go into 03 or even 40 and it can write a little bit of python like you experienced which can have mixed results but is sometimes extremely useful.

22:10

Um, you have codecs that can actually plug into a GitHub repo and and go and write some code and fix bugs and whatnot.

22:16

And then you have windsurf which can sit in your IDE alongside and be kind of that groundup adoption.

22:23

And so there's one mental model where there's like the consumers are going to ask questions and they don't even know that they should be thinking about writing code. Code will be written.

22:33

They might never see that code if they don't unfold the reasoning tokens.

22:37

Um, on the flip side, you might have someone who's using codecs to change bugs and make small changes in a GitHub repo, uh, without really opening up an IDE constantly, but they're just they're just interacting with code, but they're aware of what should be done, what should be built.

22:52

what should be built. Uh then you have a product like Windsurf where for basically a full-time programmer they're they're working in code constantly and they're solving problems at a much higher level than any of the agents can actually do but they are enhanced and sped up by these you know AI AI idees that are speeding up their their workflows and then potentially you have

23:14

this top-down enterprise AI which is what Scott's building with cognition where you know there isn't as much of a ground swell around um Devon as a consumer tool, but as I understand it uh um Devon is something that's been pushed top down on big corporations as something that is that needs to be rolled out and it has much more of a uh like an enterprise sales uh like go to market and and sales cycle. So, uh,

23:44

So, uh, interesting to see if that if that model holds of like four distinct markets like um consumer, proumer, groundup enterprise, top down enterprise.

23:56

I don't know if that's the right way to think about it, but I think that's something we'll be digging into over the over the next couple days.

24:02

Yeah, I mean ultimately some combination of productled and enterprise is probably the winning combo long term.

24:07

They all seem to be getting adoption and printing money.

24:11

the the questions just yeah churn and and and how the market looks in the really long term.

24:15

It is interesting because like AI coding tools it does feel like it's a massive market.

24:19

You looked at the numbers for uh how much money spent on software engineering. It's huge.

24:25

And then it's also just a completely new market in the sense that there isn't anything that's like really established.

24:32

Like they're not really displacing spend on anything else.

24:33

It's all just incremental and additive.

24:35

So, it's kind of hard to hard to handicap exactly what you're going after.

24:40

Um, I thought this post from Caleb Harris was an interesting call out.

24:44

He says, "Last night I dropped 12 backlog bugs from linear into codeex and launched.

24:51

Graptile found some nits in the reviews, but they were almost entirely correct and then spur QA tested almost entirely human out of the loop at this point." That's awesome.

24:58

And uh yeah, well, good time to tell you about Linear.

25:02

Linear is a purpose-built tool for planning and building products.

25:05

meet the system for modern software development, streamline issues, projects and product road mapaps.

25:10

Um anyway, thank you John.

25:10

Uh they also launched uh co-pilot tuning which allows companies to fine-tune their models with their own data to create domain specific agents.

25:19

Windows AI foundry launches as a local platform on Windows and Mac for training and fine-tuning and running LLMs with Foundry local.

25:28

Azure AI foundry models adds Groc 3 which we talked about.

25:30

MCP gets firstparty support.

25:33

Windows Subsystem for Linux, a bunch of other stuff.

25:37

Uh, SQL Server 2025 hits public preview. Let's go, SQL Server.

25:43

Huge for the database engineers in the crowd. Let's give it up. Let's give it up.

25:48

Um, anyway, the the I think the big definitive analysis of the the post game on Microsoft Build uh came from Ben Thompson, but Derek Thompson, no relation, I believe, but maybe there's some brothers.

26:01

Uh Derek, that would be wild.

26:04

He wrote uh Abundance with Ezra Klein.

26:06

Um he says, "Fascinating.

26:08

Ben Thompson's uh vision of the future of the internet, i. e. the agentic web.

26:13

Scenario is before I go to sleep.

26:16

I tell Chachi PD plan my 5-year-old's birthday next Saturday. Budget $500.

26:19

When you've made the reservation, email these two these 20 people a printable invitation to attend.

26:26

Also, my wife wants to go to England in mid July.

26:28

Find five plausible flights for the family and make several distinct itineraries.

26:31

Finally, please edit this work memo.

26:32

When I go to sleep, the AI agent negotiates slots with two bowling alleys, buys a cake, emails, printable invites, plans the trip, copy edits, etc.

26:41

This presents an interesting economic challenge.

26:43

What happens to ad revenue when more traffic is just AI?

26:47

Ads make little sense when the reader is a robot.

26:49

Well, maybe instead sites ask agents to make tiny payments that are fractions of a cent every time they call up an article.

26:56

Ben suggests the use of no fee stable coins, blockchain dollars, but I think microtransactions could work with old-fashioned dollars.

27:03

I have no idea if this cashes out, but very interesting vision of how AI would necessarily transform internet economies as traffic shifts to from humans to agents. Uh, interesting.

27:14

So, I want to I want to dig through this uh strategy article.

27:17

Um, but I also think this is significant because so every business, every consumer tech company eventually becomes an advertising business from Instacart to Uber to Apple. Yep.

27:27

And if you remove these ad businesses, it typically will just rip out a lot of the actual uh earnings, right?

27:38

Because advertising tends to be very high margin.

27:43

And so what what I think we're going to see is ultimately this kind of drawn out war between agentic software and the old internet where the old internet doesn't want to let go of advertising, right?

27:58

And so I think that um you know microract transactions are are potentially somewhere Yeah. that you could land.

28:05

But it's interesting to dig through.

28:07

I'm not entirely sure that language models and AI agents will be entirely resistant to advertising because I I imagine No, but it's more so if you're Uber Y and you have a great ad ads business. Yep.

28:19

Um are you thrilled to have an agent trolling and and booking things and buying groceries and all this stuff? You're not right.

28:28

You're like, I would prefer unless unless the ads that you run on your platform influence the future training of those AI models.

28:36

So, for example, if an LLM is training on transcript of this show right now and then we start talking about how RAMP can ta save time and money. Well, it does, John. Yeah. Yeah. Exactly.

28:48

We're we're we're stating the facts here. Yeah.

28:50

And we're explaining that RAMP has easy to use corporate cards, bill payments, accounting, and a whole lot more all in one place.

28:54

Not even to mention RAM travel. Yeah.

28:56

We didn't even get to mention RAM travel, which makes travel so seamless and and stream.

29:03

We use it every time we hit the road. We do.

29:04

Um, and so an LLM as it's trolling through the internet, it comes across a, you know, a host red ad like that that influences the future recommendations it makes, right?

29:15

It gets baked in the training data.

29:16

So perhaps ads could be just as as as influential in the sense that, you know, LLMs are simulating human behavior.

29:25

Everyone likes to say that advertising doesn't work on them, but time and time again data shows that it does.

29:31

And so if you have an agent that's actually perfectly simulating a human, it should also be simulating receptivity to advertising.

29:41

You would think that it would. That's right.

29:42

Um but uh but let's dig into Ben's article because I think I agree and I think you agree more than we're we're letting on right now, but let's dig into it.

29:52

So he starts with this with this uh 2014 article in the Atlantic by Ethan Zuckerman uh about like the death of the web.

29:59

I remember this uh this iconic article.

30:01

He says uh from Ethan Zuckermanman, I have come to believe that advertising is the original sin of the web, the fallen state of our internet. Is it a direct?

30:13

Nobody's taking this serious. Taking this seriously. Yeah.

30:15

I forget what what Jordy has.

30:18

Uh Jordy has a laugh track now.

30:21

It's called The Internet's Original Sin.

30:23

Uh and uh keep going, John.

30:23

You're gonna make this very hard.

30:28

The fallen state of our internet is a direct, if unintentional, consequence of choosing advertising as the default model to support online content and services.

30:39

Through successive rounds of innovation and investor storytime, we've trained internet users to expect that everything they say and do online will be aggregated into profiles which they cannot review, challenge, or change that shape both what uh both what ads and what content they see.

30:53

Outrage over experimental manipulation of these profiles by social networks and dating companies has led to heated debates amongst the technology savvy the technologically savvy but hasn't shrunk the user bases of these services as users now accept that this sort of manipulation is an integral part of the online experience.

31:13

Mark Andre hit the phone.

31:17

Mark Andre Mark Andre who was there when the web was born explained in a 2019 19 podcast why this sin was committed.

31:26

The quote is edited lightly for clarity.

31:28

One would think the most obvious thing to do would be building in a in the browser the ability to actually spend money. Right?

31:35

You'll notice that didn't happen.

31:37

And in a lot of ways we didn't even think it's unusual that it didn't happen because maybe that shouldn't have happened.

31:43

I think the original sin was we couldn't actually build economics which is to say money into the core of the internet and there and so therefore advertising became the primary business model.

31:53

We tried very hard to build payments into the browser. It was not possible. We made a huge mistake.

31:58

We tried to work with the banks and we tried to work with the credit card companies.

32:01

It was sort of this classic single point of failure bottleneck or at least in this case two points of failure.

32:06

Visa and Mastercard essentially had a duopoly at the time and so they were just literally if they didn't want you to be in the switch, they didn't want you to be able to do transactions, you simply weren't going to do it.

32:18

to do it. And what's interesting is like I feel like the modern entrepreneur like the the Sam Alman style dealmaker if you transported them back in time you put the you can just do things mindset in there you would just be like oh yeah

32:32

like uh the whoever's building the browser did some crazy deal with Visa and is bundling these transactions and so yeah they're totally waving the 3% fee but they're aggregating it this way and they're also investing and one watches the other. It's interesting that

32:43

It's interesting that the credit card companies at the time too, you can imagine I would pay an incremental fee to have my credit card data perfectly stored in my browser.

32:56

Like you could capture BIPS on that pretty easily. Totally. Totally.

32:58

And so you think there would have been a deal to be done.

33:00

But remember this was at a time when people were scared of the internet.

33:06

People generally weren't even fully convicted at the time that it was it it still had non-believers, right?

33:12

I wonder if it was more of a failure of imagination from Visa and Mastercard executives or a failure of aggression in the deal making of the web 1.

33:23

0 entrepreneurs because the web generation we're in right now.

33:30

You're basically saying entrepreneurs today are built different.

33:33

It's not just that they're built different is that they're very willing to go and do crazy deals with huge companies. Yeah.

33:40

Like I mean think about that's real one thing that's real back then there was more there was more almost technical risk to something to something like this that didn't happen but yeah yeah I mean a lot of the maturity of the people take startups seriously.

33:58

Um, but I'm thinking of that uh what what is that Anker Jane company where you can pay your rent in with credit card and Wells Fargo like totally took a bath on the deal at least in the short term. You remember this?

34:09

It was some sort of credit card built built. Yeah. Built rewards.

34:13

And so like doing a deal with Wells Fargo at that scale as like a startup is crazy, but like Anker is clearly just a really really great like dealmaker.

34:23

And so he was able to get it done.

34:26

And if if Points on Housing actually does plan like like play out to be a real thing in a big business, like it will be on the back of that crazy deal that needed to get done to make this happen.

34:38

Like it wasn't just going to be some groundup uh small company deal.

34:43

like you had to go in and do this this big thing. Yeah.

34:46

It's interesting in some ways.

34:49

Apple Pay enables what would have been the goal at this time, but it does it effectively at the hardware layer. Yeah. Yeah. Yeah. Yeah. It's not even that good.

34:58

I think it's pretty it's getting to the point where it's pretty good. Yeah. Yeah. It's pretty good.

35:01

But I mean, still like you you hit web checkout for all the time and have to hit the autofill, which is like such a cluge because it could just not even be there. Yeah.

35:11

Yeah, it could just be as native as you click a link or, you know, even just like the integration of email or the integration of the the share sheet.

35:20

Like there's so many different uh aspects that are built into the web browser natively like the ability to play video on the web.

35:26

Like that was something that all the browsers needed to figure out and like it just works now.

35:32

Like every browser supports video.

35:34

Uh Mark Anderson was talking about um all the different things like like why the hyperlinks are blue.

35:39

like he's the one that decided that because he was like we need hyperlinks so we we should make them blue and it's like there could have been a version of money.

35:45

So anyway, uh Ben Thompson goes on to say I think Andre is too hard on himself and I think Zuckermanman is too harsh on the model Andre created uh on the model um Andre created the conditions for the original web was the human web and advertising was one of the best possible ways to monetize the scarce resource in digital human attention.

36:07

The incentives all align.

36:10

Users get access to vastly to a vastly larger amount of content and services because they're free, which I really like.

36:16

It's great that anyone can access the internet for basically free and they can go all over and most of these services are free.

36:22

The information spreads very quickly.

36:24

Content makers get to reach the largest the information doesn't spread quickly always pay but there's the black market of PDFs as well that I think it's Paula has talked about. Yeah. Yeah. Definitely in SV. Yeah.

36:39

Um, content makers get to reach the largest possible audience because the access is free and advertisers have the opportunity to find customers they would never have been able to reach otherwise.

36:48

Yes, there are downsides to advertising.

36:50

Zuckermanman fredded about, but everything is a trade-off and the particular set of trade-offs that led to advertising the advertising centric web were on balance a win-win-win that generated an astronomical amount of economic value.

37:03

Yeah, it's hard to go back and play play out the Google story and say like, "Oh, they left a lot on the table." Yeah.

37:08

In so many ways, content content wants to be free, right?

37:13

The second that you put web pages or anything behind a payw wall, Yeah.

37:14

the uh attention they get drops by 99%. Usually more. Right. Totally.

37:22

Uh and yeah, it's just it's just way better to be advertising le.

37:27

It's way better to just, you know, as you're talking about one story, just just start talking about Figma.

37:33

You know, think bigger, design, build faster.

37:35

Figma helps design and development teams build great products together.

37:39

You can get started for free.

37:40

Anyway, it is the backbone.

37:40

It's the back of our show and we can't thank them enough. Okay.

37:46

Moreover, I disagree with Andre that we could have ended up with a better system if the banks and credit card companies had been willing to play ball.

37:55

In fact, over the last 30 years, the credit card companies have in particular have in part thanks to companies like Stripe gotten their digital acts together and are integral to a huge amount of web-based commerce, which itself is driven through digital advertising, the largest category of advertising for both Google and Meta.

38:11

That too is human in the biggest in that the biggest outcome of digital advertising is physical products and real world experiences like travel.

38:18

Digital products like apps and games, meanwhile, are themselves pursuing human attention.

38:22

What was not viable in the 1990s, nor at any time since then, was something like microtransactions for content.

38:28

One obvious problem is that the fee structure of credit cards don't allow for very small transactions.

38:34

Another problem is that costs to pro uh to produce content are front-loaded, and the potential payoff is both backloaded and unpredictable, making it impossible to make a living.

38:44

The biggest problem of all, however, is that microtransactions are antihuman, forcing a potential content consumer to continu continually decide on whether or not to pay for a piece of content is alienating, particularly when plenty of alternatives for their scarce attention exist.

38:57

I I really believe that a large amount of the Substack economy is effectively crowdfunding somebody's ability to just nerd out on a specific set of topics. I agree.

39:11

And I think it's beautiful. I agree.

39:12

Yeah, it does seem like if there's Yeah, sorry.

39:14

Yeah, I was just going to say, but a lot of the content creators that are paywalled on Substack would be bigger and more influential if it was just completely free. Yeah, right. Yeah. Yeah.

39:24

There's a big question in my mind of like where does investigative journalism go in the era of going direct and modern tech media?

39:34

Uh like where is tech Seymour Hirs?

39:36

Like you need someone who can just be like Seymour Hirs was just on the payroll of big newspapers for a long time and then could go and like hang out at a bar and talk to former soldiers about what they did in Vietnam and then figure out that the Milai massacre happened and like uncover this like massive scandal.

39:56

And we see this in Silicon Valley where there are companies that are like, you're not going to be able to put a bunch of sponsored content on a takedown of some company that's defrauding their investors.

40:08

The the win case is the John Keroo story with Theronos where he was able to be on the Wall Street Journal payroll for a long time, hunt that story down for a long time, do a lot of sourcing and a lot of confirmation because it's extremely risky to write something negative about a10 billion dollar company that has incredible resources.

40:25

I mean the hybrid you kind of need the heat shield oft a big organization. Totally. Yeah.

40:29

And also just you need you need to be able to pay your mortgage while you're doing the research and not publishing and and just doing nothing basically.

40:36

Uh and and being 100% focused on on chasing down a really big story.

40:42

And then when you do it there's a huge outcome because he of course sold a book and then the rights to the documentary and the movie and the TV show and all this stuff.

40:51

And I I don't even know if he's I don't know what he's up to now, but it seems like worked out really well.

40:56

Yeah, it'd be great to get him on the show.

40:58

Um here if he's getting back in the game, digging into some of the the maybe I mean these 2. 0 is coming right now. He's like, I'm back.

41:06

He's like, I'm I'm I'm ready to start talking to the employees at this new the new blood testing company.

41:11

It would have been huge missed opportunity to not just name the company Theronos 2. 0, Inc. Yeah, it' be great.

41:18

I mean, do do we know what the name is yet?

41:20

I don't know if I missed it, but we'll have to look it up.

41:24

Anyway, subscriptions do work at smaller scales, says Ben Thompson, because they are ultimately not about paying for content, but giving money to another human, which you've mentioned, or human institution from the local news business model.

41:35

Uh, which Ben wrote, it is very important to clearly define what a subscription means.

41:40

First, it's not a donation.

41:42

It is asking a customer to pay for money for a product.

41:44

What then is the product?

41:45

It is not in fact any one article.

41:48

A point that is missed by the misguided focus on microtransactions.

41:52

Rather, a subscriber is paying for the regularly regular delivery of well-defined value.

41:56

Each of those words is meaning is meaningful.

41:59

Paying a subscription is an ongoing commitment to the production of content, not a one-off payment for a one-off one piece of content that catches the eye. Regular delivery.

42:08

A subscriber does not need to depend on the random discovery of content.

42:12

said content can be delivered to the subscriber directly, whether it be email, a bookmark, or an app. Well-defined value.

42:17

A subscriber needs to know what they are paying for.

42:19

The last point is the crux of why many ad-based newspapers will find it all but impossible to switch to a real subscription business model.

42:27

When asking people to pay, quality matters far more than quality, and the ratio matters.

42:30

A publication with one valuable article a day about a well- definfined topic will easily earn more subscriptions than one with three valuable articles but 20 worthless ones covering a variety of subjects.

42:43

Yet all too many local newspapers built for ad-based business model that calls for daily content to wrap ads around spend their limited resources churning out daily filler even if those ads no longer exist.

42:54

Uh Ben says, "I expect that this model will endure in the age of AI.

42:59

Obviously, I'm biased at this point, but in a world of infinite content on demand, common content becomes community.

43:05

And if I'm successful, this essay will generate a lot of discussion amongst a lot of people precisely because it is both original and widely accessible, funded by an audience that wants me to keep writing articles exactly like this.

43:17

So, of course, Ben writes free articles and then also private adup supported articles and goes back and forth.

43:22

So, he gets a little bit of the reach.

43:24

So he goes in here, the death of the adsupported web.

43:26

The ads supported web, particularly textbased sites, is going to fare considerably worse.

43:31

In fact, the most substantive push back to my defensive advertising was in my own excerpt.

43:37

Most adup supported content is already terrible thanks to the bad incentives both Zuckerman and Andre bean and the impossible economics enabled by zero marginal cost content generation and consumption.

43:47

and consumption. Google in its most idealized form aggregated content consumers aggregated content consumers by mastering discovery in this world of abundance directing users to exactly the site they were looking for which was monetized through ads that were sold and

44:03

served by Google indeed this is the great irony in the ads antitrust case in which Google is currently embroiled asked on mobile dub memo I've heard arguments that because Google suppressed competition in open web advertising markets those markets should flourish when Google's monopoly is broken. But my

44:19

But my sense is that this ignores two realities.

44:22

First, the consu that consumer engagement has shifted into apps and walled gardens irreversibly.

44:25

Of course, like most people get their news from X or Instagram or Facebook. Um, yeah.

44:32

And this was the logic for X to say, "Yeah, we don't want you posting links.

44:36

fine for you to post an excerpt or a little bit higher because we want you to click into the content and see another ad and then click out, go see something else, see another ad and just do that forever.

44:47

And second, that Google was keeping the open web on life support and the open web's demise will be hastened when Google no longer has an incentive to support it.

44:56

What happens to the open web when its biggest, although albeit imperfect benefactor loses the motivation to sustain it?

45:02

Well, everyone's going in the walled garden.

45:06

Wald gardens like social networks are both more attractive to most users and also better for advertisers.

45:11

Google might soon lo lose what little motivation they had left to support the open web.

45:16

However, that's not Google's and the web's only problem.

45:20

Why go through the hassle of typing a search term and choosing the best link, particularly as search results are polluted by an increasingly overwhelming amount of SEO spam now augmented by a generative AI when chatbt or Google itself will simply just give you the answer that you're looking for.

45:33

In short, every leg of the stool that supported the open web is at best wobbly.

45:38

Users are less likely to go to adup supported contentbased websites even as the long tale of advertisers might soon lose their conduit to place ads on those websites leaving said websites even less viable than they are today and they're barely hanging on as it is.

45:53

I mean I can you think of the classic example of this is like historically trying to find an ingredient like a a recipe, right?

46:01

You're like I want to make oatmeal. Yeah.

46:03

And the site like buries the only information you actually need, which is like how much how many how many cups of oatmeal to water, the ratio or whatever. Ridiculous.

46:12

And you have to see you have to scroll by, you know, two popups and six ads to get to that little nugget of information.

46:19

And now it can just be immediately surfaced by an LLM.

46:21

I'm trying to think about the last time I visited like a truly adupported website.

46:29

Like I'm scrolling through here and it's just it's just nothing in my tabs.

46:32

Like everything is either a like a a marketplace that monetizes by transaction fees or a paywalled website for news or just Google search generally finding stuff.

46:52

But like there are very very few websites that I visit today on a regular basis that are purely adupported.

46:59

Like I do go to the Wall Street Journal or the Economist or um strate or any of these but they none of them are purely ad supported.

47:07

The Wall Street Journal does have ads but uh they also have a big payw wall and a and an actual subscription fee. So it's interesting.

47:15

Yeah, the open web I mean is it is it dying or is it already dead?

47:20

like it seems like like no like like you know the the the Vice News and the Buzzfeed like that boom and bust has already happened. Yeah. So I don't know.

47:29

Uh anyway, he moves on to I mean what do you think?

47:31

It'd be interesting to try to honestly run a deep research report on trying to understand traffic to these websites that are effectively dying but not dead, right?

47:42

Because if you have a website that's I would imagine like how many websites have recipes for uh you know cooking various things that still get people trickling in and they're technically still monetizing. The question Yeah.

47:57

The question is like at what point does uh does it cost more to even host the website than it's generating traffic, right? Yeah.

48:04

And there's been a couple of those roll because if you're making whatever if you're making $1,000 a month in ads and it costs you a couple hundred bucks to maintain the website in theory somebody would just keep that running uh indefinitely. Yeah.

48:21

I mean uh Techrunch was obviously famously ad supported.

48:23

They were bought by Regent uh which is a private equity rollup of uh of brands.

48:29

Let's look at their portfolio.

48:32

It's all very hyper specific stuff.

48:34

They own Techrunch now, but they also own Computer World, Network World, PC World, Mac World, TechHive. They own Cheddar News.

48:41

They own Defense News, Air Force Times, Navy Times, Army Times.

48:47

So, I imagine that Navy Times is probably ad supported.

48:50

I mean, I'm just clicking around on the website and there's not a lot of like pay walls I'm hitting.

48:55

pay walls I'm hitting. um but they are running ads and I'm sure that they have other ways to monetize whether it's through uh premium you know um reports or maybe conferences or something like that but in general it seems like there aren't that many companies I I would struggle to name companies that are generating over a billion dollars in revenue off of purely adbased website

49:17

that's not an ad exchange for example yeah right doubt it exists it doesn't really exist anyway uh so Ben takes all of this I uh all this like preliminary thesis and and weaves it into Microsoft and the open agentic web says this reality is the fly in the ointment of an intriguing set of proposals that Microsoft put forward yesterday at the build 2025 developer conference about the open agentic web. Um and so the

49:42

Um and so the quote from the CTO is the thing that is super important if you think about what an open agentic web could be is you need agents to be able to take actions on your behalf.

49:54

And one of the really important things about agents being able to take actions on your behalf is that they have to be plumbed up to the greater world.

50:02

So you need protocols, things like MCP and A2A and things that are likely going to be emerging over the coming year that will help connect in an open, reliable, interoperable way.

50:10

the way agents that you are writing and agents that are being used so actively now by hundreds of millions of people uh to be able to go access content to access services to take action on behalf of users in fulfilling the tasks that have been delegated to them.

50:25

One aspect of this vision of the agentic web says Ben Thompson was Microsoft's commitment to MCP which we talked about a little bit created by anthropic.

50:33

Scott told Neil Patel in an excellent interview on The Verge that while MCP wasn't exactly what he would have designed from scratch, ubiquity is more important than semantic differences, particularly when you're trying to create HTTP for AI agents.

50:47

The second part of Scott's vision was something Microsoft created called NL web, natural language web, and these are natural language interfaces for websites that make them more directly accessible for agents.

50:56

He says, "If you think back to the web, we have HTTP and then we had things that sit on top of HTTP like HTML mainly that are opinionated about the payload.

51:06

So today we're announcing an NL web.

51:08

The idea behind NL web is that it is a way for anyone who has a website or an API already to very easily make their website or API an agentic application.

51:20

Lets you implement and leverage the full power of language large language models to enrich the services and products that you're already offering.

51:26

And because NL web endpoint the every NL web endpoint is by default an MCP server, it means that those that those things that people are offering up via NL web will be accessible to any agent that speaks MCP.

51:40

So you really can think about it a little bit like HTML for the agentic web.

51:44

I always thought this was odd because I thought that the that the layer to go between APIs and HTML for an agent to just transform HTML into something that an agent could understand would be very very simple.

51:54

I didn't think that we'd need a new protocol, but it seems like it's getting tons of adoption, so I guess this makes sense.

52:00

Um, we have done a bunch of work already with our partners who have been really excited to be able to very quickly get to implementation and prototypes using NL Web.

52:08

They've worked with Trip Adviser, O'Reilly Media, a ton of really great companies that offer important products.

52:12

So, Scott concluded by reemphasizing how important it was that layers that the layers of the agentic web be open and use the evolution of the internet as his reason of why.

52:21

So he says, "So the last thing that I want to say before handing things back over to Satia is just sort of press on these two points about why open is so important here.

52:33

You know, it is unbelievable what can happen in the world when simple components and simple protocols that are composable with one another are out there exposed to the full scrutiny and creativity of every developer in the world who wants to participate or who has an idea.

52:46

This thought game that I play with myself all the time is trying to imagine what the web would have looked like if one of the actors in the early development of the web say the browser manufacturers.

52:55

Browser manufacturers is a great terminology called Mark Andre a browser manufacturer.

53:01

Yeah, he made his money in in browser manufacturing.

53:06

Domestic browser manufacturing.

53:06

Domestic browser manufacturing.

53:07

uh American Dynamism browser.

53:12

So had decided that they wanted to vertically integrate and own the entire web.

53:15

A 100% of the web would have been dictated by the limits of their imagination.

53:19

And it's just obvious with 30 years of history that that wouldn't have been a very interesting web.

53:22

The web is interesting because yeah, think of of the iPhone app layer.

53:26

If Apple wasn't controlling, it would be extremely chaotic, but there would undoubtedly be sort of cool uh, you know, things that would emerge.

53:37

Yeah, I'm kind of split on it because like there's so much creativity that you can do within a JPEG or an MP4 and so if you get on YouTube, I mean it requires more production value, but uh you know we we we like we are not constrained by anything other than just pixels here and we can put tickers and do all sorts of things within the confines of the MP4 box.

54:00

I guess it is a little bit more narrow than than a web UI that anyone can use and click around.

54:06

can use and click around. So we in some ways we are being constrained but there like you just scroll on on any social network and you see the amount of creativity that that occurs even within the constraints of like no links right yeah despite Apple's control of the app store you still get the benefits of the web which he goes on to describe here

54:26

the web is interesting because millions tens of millions hundreds of millions of people are participating to make it this rich dynamic thing that's what we think we need with the gentic web and that's what we're hoping you all can get get inspired to go work on a little bit to riff on and to use the full extent of your imagination to help make this thing interesting. Yeah. So he says uh so back Yeah.

54:43

So he says uh so back to Ben Thompson he says I think the widespread adoption of MCP as a protocol layer and NL web as markup layer sounds excellent.

54:52

The big hole in Scott's proposal however was pointed out by Neil Patel in that interview.

54:55

Uh that's the piece that on the web right now seems most under threat.

54:59

the underlying business dynamics of if I start a website, put a bunch of schema on it that allows search engines to read my website and surface content across different distributions, I might add an RSS feed, which is a standardized distribution everyone agrees on.

55:11

There's lots of ways to do this, but if I wait, if I make a website and I open myself up to different surfaces, what will I get in return for that is not necessarily money.

55:20

In almost every case, not money.

55:22

What I'll get is visitors to my website and I'll monetize them however I choose, selling a subscription, display ads, whatever. That's broken, right?

55:27

As more and more of the answers appear directly particularly in AI based search products traffic to website has websites has generally dropped.

55:35

We see this over and over again.

55:36

What's going to replace that in the agentic era where we've created a new schema for agents to come and talk to my website and receive some answers.

55:44

What's going to make it worth it?

55:44

And this is a good question.

55:46

Uh Scott in his answer noted that websites would be able to communicate to agents that that they wanted to make available and on what terms along with some vague handwaving about new advertising models and transactions.

55:58

That last point is valid.

56:00

Trip Adviser sells hotel rooms.

56:00

O'Reilly sells training courses.

56:02

And you can see a world where websites based on transactions can not only benefit from exposing themselves to agents, but in fact transact more and potentially pay an affiliate fee. Patel.

56:11

So yeah, I mean if you're if you're searching for a car and you're searching across eBay and Autotempest and Cars and Bids and Bring a Trailer, uh all of a sudden it doesn't really matter if someone's landing on your website.

56:26

All that matters is that they're transaction transacting and that they found the car that they're looking for.

56:30

And so for all those transactionbased websites, they're going to do fine. Yes.

56:35

Someone who's just making content and putting ads on it is kind of screwed.

56:40

I don't know about these transactionbased websites, right?

56:41

I mean thinking of cars as an example, right?

56:45

You have car gurus, you have probably what cars.

56:48

com, bring a trailer, cars and business auction ones.

56:51

Imagine a world where I just have car gurus is like an adbased business model.

56:57

They want to drive leads to dealerships and get that.

57:02

But in a world where I just say, "Hey, find me every single Ford GT between this year and this year that's publicly available online.

57:08

bring me the contact information and actually reach out to the people. Yes. Yes. Yes.

57:13

But but you think you I don't think you can disintermediate entirely.

57:17

I think that I I think that the agent would would say, "Hey, there's a Ford GT that's available on Bring a Trailer right now.

57:23

Here's the current price.

57:25

Do you want me to sit here and bid on it for you?" Yeah.

57:26

But here's the issue though because I can have the you could have the agent like in theory the agent would go to CarGurus and fill out the form and tell the dealer process a lead to the dealer and the dealer's like thank you for this lead car gurus if I sell the car I will pay you a fee or maybe it's just paid up front.

57:44

But the issue is an agent could theoretically just like find the see the image of the car and then find do an image search. Uh maybe. Yeah.

57:51

I I'm just saying do an image search and be like, "Oh, I found this exact car and this exact photo that the dealer posted on their own site, their own website, and I'm not going to pay carers.

58:01

I'm just gonna find and I'll actually call the dealer and confirm that they have it and I'll negotiate on your behalf and then but people already do that. Carers is cooked." Yeah. Yeah.

58:10

They already do that, but it's just much more scalable to say Yeah.

58:14

But I mean that's not that's not a a marketplace like like if you are on bring a trailer and you see a car like you generally can't disintermediate bring a trailer. Sure.

58:24

Like like so if you are a platform that's that's bringing liquidity to the market and bringing listings that don't exist elsewhere on the internet.

58:30

But yes, I mean people see this with Airbnb where they they they they see a beautiful place and they're like I'm going to go just look up this place and see if they have their own website.

58:39

But a lot of people that are selling a one-off car don't have a website where they're listing their car.

58:44

Like somebody might do that.

58:47

Maybe in the future you would just you would just, you know, post on X, hey, I'm selling my car.

58:50

I have it listed on Bring a Trailer, but you can just negotiate. Well, it's funny.

58:55

So, one of the one of the companies that's probably pretty cooked in this new era is internet brands.

58:58

and they own WebMD, they own but all the way through they own Ford truck enthusiasts, they own Ren list, they own they own the Dodge forum.

59:10

Like some of these more community based sites I think are going to be fine. Yeah. Yeah.

59:13

Because you got because it's a place where discussion is happening, but again uh Oh well well sites that are purely indexing seem Yeah.

59:25

Ben Thompson goes on and says, "The original sin of the internet lacking native payments was not, in my opinion, a sin at all.

59:30

Advertising supported the human web.

59:31

Not because Andre failed to make a deal with the credit card companies, but because it was the only business model that made sense.

59:40

No, the real neglect and missed opportunity in terms of payments is happening right now.

59:44

Microsoft is on to the right idea with its adoption of MCP and in an introduction of NL web.

59:48

But its proposal by virtue of not including native payments isn't nearly compelling as it should be uh as compel as compelling as it should be.

59:58

The key difference from the 1990s is that on the agentic web native digital payments both are both viable and the best possible way to not only keep the web alive but also in the process create better and more useful AI.

1:00:11

And so he goes on to talk about stable coins and agentic microtransactions.

1:00:16

Let's start with the viability. This is from Bloomberg.

1:00:20

Stablecoin legislation overcame a procedural blockade in the US Senate, marking a major victory for the crypto industry after a group of Democrats dropped their opposition Monday.

1:00:28

The industrybacked regulatory bill is now set for a debate on the Senate floor with a bipartisan group hoping to pass it as soon as this week, although senators said a final vote could slip until after the Memorial Day recess.

1:00:41

Hopefully this passes and we can talk to everyone on crypto day about this uh next week.

1:00:44

That'll be a lot of interest, a lot of fun.

1:00:46

Um Ben goes on to say, "I know I've driven uh longtime strategy readers a bit baddy with my long running and still enduring in the face of massive grift and seemingly unending scandals interest in crypto, but stable coins are genuinely a big deal."

1:00:59

I wrote a brief explainer last fall when Stripe acquired Bridge.

1:01:02

And so I'm sure most people are familiar with stable coins.

1:01:05

Um, he goes on to say, "Stable coins solve several of the microtransaction problems I listed above, including dramatically lower or no fees and the fact that they are infinitely divisible and thus can scale to very small amounts.

1:01:18

Stable coins, by virtue of being programmable, are also well suited to agents.

1:01:22

Agents, meanwhile, are much more suited to microtransactions because they are not in the end simply software making a decision because they are in the end simply software making a decision unencumbered by the very human feeling of decision paralysis.

1:01:32

So an agent can just go around and make a bunch of microtransactions.

1:01:36

Um the entire digital ad ecosystem, he writes, is an example of deterministic agents making microtransactions at scale.

1:01:42

Every time a human loads a web page, an awe inspiring amount of computation and communication happens in millisecond in milliseconds as an auction is run to fill the inventory on that page with an ad that is likely to appeal to the human.

1:01:56

These microtransactions are only worth fractions of a penny, but the aggregate volume of them drives trillions of dollars worth of value.

1:02:01

The problem, as I and Patel both noted, is that this ecosystem depends on humans seeing those web pages, not impersonal agents impervious to advertising, which destroys the economics of adup supported content sites, which in the long run dries up the supply of new content for AI.

1:02:21

Open AAI and Google in particular are clumsily addressing the supply issue by cutting deals with news providers and user generated content sites like Reddit.

1:02:29

This however is bad for the sort of competition Microsoft wants to engender and ultimately won't scale to the amount of new content that needs to be generated.

1:02:36

What is possible, not probable, but at least possible is to in the long run build an entirely new marketplace for content that results in a new win-winwin equilibrium.

1:02:43

First, the protocol layer should have a mechanism for payments via digital currency, stable coins.

1:02:49

Uh second, providers like OpenAI should build uh an auction mechanism that pays out content sources based on the frequency with which they are cited. AI answers.

1:02:58

And this is kind of already happening with Spotify, right?

1:03:01

Like there's a whole bunch of music on Spotify.

1:03:02

They have a whole pool of capital from all their subscribers and then the more listens that they get, the more money flows in.

1:03:07

And it's not really an ad-based auction system.

1:03:09

It's more like just dividing up the pie.

1:03:12

The thing there, and it's a good comp, is that that was happening from the very very beginning in the initial deals that Spotify did with record labels to get the music on in the first place.

1:03:25

And now we're in a situation where all the LLMs are just using the internet. That's not true though.

1:03:32

Like OpenAI has a deal with the Wall Street Journal and they have some deals, but they certainly don't have deals with every book they've ever ingested. No. No. No.

1:03:41

But you imagine that they'll kind of go down the power law curve of like the most important information sources and broker deals more and more and more, which is what Spotify did at the beginning because if they're sued, remember like they had to and they have been sued.

1:03:55

They've been sued a ton and then over and over.

1:03:56

So, so they get sued and then they settle and then they do a deal and like yeah, it's a little messy but it's fine because at the end of the day the Wall Street Journal has a deal with OpenAI where OpenAI can ingest information from this surface it and chat GPT results and the Wall Street Journal gets paid for that work.

1:04:12

And and I don't know the structure of the deal, but I imagine that it's something like like, you know, we're generating this much money from people searching and and you're driving 1% of our value and so we'll give you 1% of the cut or half a percent. We'll split it with you.

1:04:28

Just like, you know, if you're Taylor Swift on Spotify and you're getting, you know, 1% of the listening time, they'll give you half a percent of revenue or something like that, right? Is that not crazy? That crazy. I love it. theoretically.

1:04:40

I think that's like what what's happening and and the problem is just going to need to go down the stack until it's self-s serve because if you want the open web, you can't do individual deals.

1:04:48

And so eventually anyone can go on Spotify. Like we're on Spotify.

1:04:52

We get uh I think we get a check.

1:04:54

We get a check from Google at least.

1:04:55

Um and in theory it's like we don't need to do a deal with Spotify.

1:04:59

Of course Joe Rogan did do a deal with Spotify directly like the Wall Street Journal did a deal with OpenAI.

1:05:05

Um but plenty of podcasts go on Spotify.

1:05:07

Spotify automatically runs ads and and does premium plays and then sends them a check.

1:05:13

It's just such an interesting dynamic, right?

1:05:14

Because why would it play that?

1:05:17

So often LLMs are not just serving the content, they're effectively remixing it. Totally.

1:05:25

And the precedent for that is like historically hitting the page.

1:05:29

Historically, if the journal went into the, you know, somebody from the journal goes into the library, public library, reads a bunch of books, forms its own kind of summary of nobody's getting paid for that, right? Yeah.

1:05:43

But and so there's an argu the the models can make an argument that there's enough precedent that says there's nothing illegal about ingesting. Yeah, that's true.

1:05:54

But there is this there is this win-winwin that Ben Thompson's talking about.

1:05:58

this this idea that uh Google has an incentive to keep the open web alive and the LLM providers might have a similar incentive because they want more journalism and more facts to hit the internet that they can scrape and pull into results.

1:06:10

And so what is the value of a single Wall Street Journal page loading logged out right I could see something scenario where it's like net new content or something like that where like OpenAI doesn't deal with Substack. every time. Why not? Why not on a user basis?

1:06:30

Uh I asked for a summary of the Wall Street Journal mansion section and it went read 10 articles.

1:06:38

I'm saying I'm saying every time.

1:06:41

I'm just saying it's possible that like there needs to be some type of cut off date where where you know again how does this actually it sounds really simple and finding this win-winwin but the actual mechanics of getting a deal done between the internet and a relatively small group of foundation model providers but again are you saying that there's a tremendous amount of complexity?

1:07:07

I'm saying there's a tremendous amount of complexity, John, because that's exactly what Ben Thompson's concluded with.

1:07:12

He said there is to be sure a tremendous amount of complexity in what I'm proposing and the path to marketplace for data generation is quite unclear at the moment.

1:07:22

Who however could have predicted exactly how the adup supported web would have evolved or centrally designed the incredible complexity that undergurs it.

1:07:29

This is where Scott's ex exhortation of openness is spoton.

1:07:31

A world of one dominant AI making business development deals with a few blessed content creators and scraping the carcass of what remains on the web for everything else is a far less interesting one than one driven by a marketplace, auctions, and aligned incentives.

1:07:47

To get there, however, means realizing that the internet's so-called original sin was in fact key to realizing the human web's potential, while the actual mistake would be in not building payments now for the coming agentic web.

1:08:01

And so bull market in stable coins, we could see stable coins trading at 1. 00001. So funny.

1:08:10

It's it's so funny because everybody's so bullish and convicted on stable coins now, but it's like how do how do we make money on this? How do we make money?

1:08:20

I don't think it's launching the next circle, right?

1:08:22

Or another stable coin provider, even though that clearly is a great business if you look at Yeah.

1:08:27

Uh, you know, some of the different providers.

1:08:30

I was actually wanting to get uh there had been rumors that Circle is talking with Coinbase and Ripple about a sale because they put out their S1, the market hated it and they're basically like you are completely dependent on on Coinbase in many ways and paying all such a large part of your revenue to Coinbase, you know, does this.

1:08:55

So anyways, could make a lot of sense for those two to join forces, but Coinbase is also dealing with a lot right now. Yeah.

1:09:06

Uh well, that concludes Microsoft and we've talked about Google a little bit.

1:09:13

Should we move over to Stargate?

1:09:15

And I would like to do that.

1:09:15

We're taking a full tour today.

1:09:18

We we're just going around the horn to all the big labs, one story after another.

1:09:21

Uh, we got to get Boots on the ground in Abalene.

1:09:26

Yeah, we got to go there.

1:09:26

I mean, Chase was was texting me earlier, we got to have him on the show, but uh I think visiting would be worth it.

1:09:33

Uh it seems like uh Sharon Gafar, I don't know if I'm pronouncing that right, but Sharon over at Bloomberg uh got an exclusive look inside Project Stargate, the 500 billion dollar AI infrastructure project, including an on the ground tour of its first massive data center in Abene, Texas, with interviews of Sam Alman. Uh love it.

1:09:54

Uh she's been on an absolute tear. Wrote this today.

1:09:57

Dropped the anthropic profile of Daario yesterday.

1:10:02

Um just banger after banger.

1:10:05

So, inside the first Stargate AI data center, OpenAI, Oracle, and SoftBank hope the site in Texas is the first of many across the United States. Let's dig in.

1:10:14

Trucks carrying concrete and electrical wiring plot over red clay weaving in between cranes and excavators.

1:10:20

Two symmetrical buildings stand in a massive plot of land where thousands of people in brightly colored vests work day and night to construct six more near identical structures that will make up the first site for Stargate project.

1:10:31

Stargate is a collaboration between OpenAI, Oracle, and SoftBank with promotional support from Donald Trump. Love it. Throwing that in.

1:10:35

Still unbelievable that they got him to just go and and announce it even though the guy loves building in America, you know.

1:10:46

This is like his whole brand.

1:10:46

Showbiz loves build the big things.

1:10:48

Uh to build data centers and other infrastructure for artificial intelligence throughout the United States.

1:10:54

The companies have pledged to spend as much as $500 billion, a number so large it's hard to believe it'll actually happen.

1:11:03

But at least for one for this one in Abalene, Texas, 180 miles west of Dallas, Chase Lock Miller says they're good for the money. Chase Lock Miller. This is an incredible. He chases the energy. He locks it down. Then it's Miller time. Chase Lock Miller, baby. Determinism is wild.

1:11:21

Let's pull up the photo of Chase. He's looking great. He's hanging out. An absolute dog.

1:11:25

Yeah, he's been on this for so long.

1:11:27

Lock Miller is the founder years ago.

1:11:29

You can go to the next slide.

1:11:31

It's so he's more centered.

1:11:33

Um I I met him four or five years ago when they were doing uh oil and gas flaring for crypto mining and it was there was and it was I mean the business was doing pretty well.

1:11:42

It's funny because it's funny because it was very low status to do the crypto to AI pivot yet you get coreweed and cruso out of it. Just absolute monsters. Always ignore the meme.

1:11:54

Whatever the meme is just ignore it.

1:11:56

build a GPT rapper and sell it to OpenAI for three billion.

1:12:00

Like pivot from crypto to AI and IPO for 40 billion. Do it. It's fine. It doesn't matter.

1:12:05

Ignore what the haters think. Yeah.

1:12:07

So Lo Miller is of course the co-founder and chief executive officer of Crusoe, a startup that helps develop AI data centers.

1:12:13

Crusoe's costs to build the one in Abalene are expected to reach 12 billion, Lock Miller says.

1:12:18

And that's not counting the billions of dollars worth of Nvidia Corp chips that will be installed in the finished facility.

1:12:24

I've never built anything of this scale.

1:12:26

Lo Miller says, "I've had to learn a lot on the job."

1:12:30

So, the three main companies behind Stargate could say the same thing.

1:12:34

OpenAI knows AI, but has relied on Microsoft for data centers.

1:12:36

Oracle knows databases, but only holds 3% of the cloud market.

1:12:41

Soft Bank once raised a hundred billion dollar fund, but that didn't go so well.

1:12:49

Let's give him some time. Let's let him cook.

1:12:50

You know, uh 10 year 10 10 year horizon.

1:12:52

We got a got a few years left on that one.

1:12:55

Oh yeah, we we we got to pivot really quickly to the to the economist because The Economist has an article about will OpenAI ever make real money.

1:13:03

We should do a whole deep dive on this one, but uh the way they describe SoftBank is hilarious.

1:13:10

Uh here it says uh they're talking about uh the latest OpenAI fundraising.

1:13:16

Uh, happily for the CFO of OpenAI, moneymen swept up in the AI mania need little persuading.

1:13:22

They are falling over themselves to fund OpenAI.

1:13:28

On May 13th, Soft Bank, a Japanese tech piggy bank, said that its $30 billion investment in the firm was unaffected by OpenAI's recent decision not to ditch its odd governance structure.

1:13:39

A nonprofit board would keep control of its forprofit arm.

1:13:44

Uh, it's a whole it's a whole great article.

1:13:46

I mean, we were talking about this off air.

1:13:48

I mean, it's absolutely credit to Sam. Yeah.

1:13:50

For not completing the conversion, but not having to renegotiate at least top level numbers, right?

1:14:00

The 30 on 330 or whatever it was.

1:14:03

Um, one of the best deal makers of all time.

1:14:07

And Masa doesn't clearly, you know, clearly doesn't care what price he pays.

1:14:13

Simultaneously, the soft bank deal is like crazy and unprecedented and wild and Japanese piggy bank and all that, but the premise of like will soft will open AAI ever make real money is like ridiculous to me.

1:14:24

Like will the next will the next uh consumer tech giant make money online?

1:14:32

Will a company 500 million users make money?

1:14:35

I was saying saying this yesterday off air.

1:14:37

It was like you remember it was only it was less than a year ago.

1:14:41

people uh Anthropic was ripping. Yep.

1:14:44

And Yahoo and everybody was just saying, "Yeah, open AI is Yahoo. Open AI is Yahoo."

1:14:48

I was thinking about this a lot and I was like, "What what was the problem with Yahoo?"

1:14:53

like they got like like they they got disrupted by Google obviously but like more specifically what happened like it was it was a shift from from non-algorithmic links right to a better algorithm in page rank like it was a development of a better algorithm that really smoked them and I'm wondering like is that is that a point of fragility for open AI like you could imagine someone comes up with a different model for AI.

1:15:25

Like they say like the transformer architecture is fundamentally not it.

1:15:30

It's V1 and we're coming out with V2 and it's completely different architecture.

1:15:35

Fine, maybe that happens.

1:15:37

Is OpenAI agile enough to implement that within six months? Like probably, right?

1:15:41

If Deep Seek can clone transformers and reasoning models in just a few months, you would think that like I guess the question is like the Yahoo failed not just because Google invented page rank which sorted the links by the number of links to a specific website.

1:15:59

And so if everyone is linking to stratey and you search for tech news that increases the page rank and that shows up at the top.

1:16:06

That was the that was the key insight by Larry Page, the creator of Page Rank.

1:16:09

And it and it created much much better results.

1:16:15

Like when you went to Google, it was dramatically better of a product.

1:16:17

I remember when it came out like I was a kid and and I remember using Alta Vista. You were what like Yeah. At that point. Yeah. Yeah. Exactly. Like little late teens.

1:16:28

I think I was actually like eight or something.

1:16:30

Um anyway, uh so so I had been using Yahoo and Alta Vista and you would search and it would just be kind of like random links or they'd be like kind of curated.

1:16:40

It would literally just be it would it would basically just search based on the keyword.

1:16:44

So if you searched for like tech news, you would more than likely just see like technews.

1:16:49

com come up because it had really strong like SEO just from keyword searching. Yeah.

1:16:54

Google brought in page rank and said, "Well, if if it doesn't matter that tracer, the name the word doesn't make any sense and Ben Thompson never actually writes IM tech news technology news technology.

1:17:04

It doesn't matter that he says that.

1:17:07

All that matters is that everyone is linking to stratey and it's actually a great source for tech news and so we're going to put that at the top of the results."

1:17:13

Now, that was a better algorithm and and it was a better product very quickly.

1:17:18

But if Yahoo had implemented page rank, which I don't know if it was like patented or something, but if they'd been able to to keep up with that, people probably would have just stuck with Yahoo for a lot longer, but something about where the company was and the culture didn't allow them to upgrade their search results.

1:17:33

And so they fell behind very quickly.

1:17:36

And so for OpenAI to become the Yahoo of AI, there would have to be a new company that comes out with a dramatically different technology or paradigm and either is able to patent it or or keep it locked up in some way that OpenAI can't just port that innovation back because they've obviously been able to do that with other advances in image models that are happening elsewhere, reasoning or tool use. Like it would be crazy.

1:18:01

And this is something that's always interesting in tech is like there just don't seem to be that many patents that hold.

1:18:06

Like how amazing would it be if you're just like I'm the person that thought of MCP or I'm the person that thought of image diffusion and I have the patent and so I'm the only one that can build the company or you have to pay me to license this.

1:18:17

Like that's just not a conversation at all.

1:18:20

How many designers, for example, have patents from big tech companies and then can leave and effectively rebuild the same feature?

1:18:28

And the patents don't even seem to hold like like did you know that uh the p swipe down to refresh, pull to refresh that was patented by a designer at Twitter like 10 years ago or 15 years ago.

1:18:40

And essentially they created the slot machine for the mind. Yeah.

1:18:43

Endless scroll was another one.

1:18:46

Slot machine for the mind. Yeah. Endless scroll.

1:18:48

Well, now now you don't even really pull to refresh.

1:18:50

I mean, I guess you do every once in a while, but mostly it's just end the scroll.

1:18:53

But that was another thing that was invented at one of the big tech companies and then immediately ported across every other platform and like stories in uh Snapchat.

1:18:59

Evans Spme Spiegel created the stories format.

1:19:05

Thank you Evan Spiegel for creating stories. We love innovation.

1:19:06

Thanks for creating a lowrisk way to just share what you're up to.

1:19:13

Really shilling for stories.

1:19:13

but he wasn't able to patent it or like or like you know keep meta from rolling it out to every single thing.

1:19:19

to every single thing. Uh LinkedIn has stories now and and there's this question is like that it's kind of this weird failure potentially of the IP system maybe it's better because stuff just goes everywhere but if you're if you're open AI and you're on top of the

1:19:32

in of the innovations that are happening even at other labs you would imagine that you'd be able to port back anything and stay on top as long as you're that front door and you have the scale that Yahoo or Google eventually got to mill hundreds of millions of users you should be fine anyway. Uh that was just my

1:19:45

Uh that was just my thinking on the the Yahoo thing. feels like a stretch. Yeah.

1:19:50

Especially given the recent uh the recent innovations. Totally.

1:19:55

Anyways, back to back to Abolene. Yeah.

1:19:57

Where we don't use Manis, we use we use chat GPT around these parts.

1:20:03

Uh anyway, so the artic the article here is basically pointing out that OpenAI, SoftBank, and Oracle Yeah.

1:20:10

are funding this $500 billion project and yet none of them have hyperscalers.

1:20:17

of them have hyperscalers. hyperscaler sort of data center Microsoft but yeah experience um no Microsoft I mean Microsoft is not in Stargate it's open AI soft bank and Oracle yes yeah true true and so the only companies that have

1:20:31

built cloud infrastructure on this scale Amazon Microsoft yeah Oracle knows databases but they only have 3% of the cloud market right oh they're saying Larry Ellison Larry Oh yeah right you're in the you're in the truth zone Larry's Good Larry SCed. Don't

1:20:46

Don't Don't come for Larry. Larry's on top of it.

1:20:51

We will defend seven years old and he looks like he's 42. He looks great. Incredible. And he follows TBPN.

1:20:56

So he's going to be fine.

1:20:58

He's he can he can build some cement.

1:21:00

He's got Chase on the case. That's right.

1:21:02

He's got Chase on the case.

1:21:04

Anyway, so in interviews with Bloomberg Business Week, the CEOs of OpenAI and SoftBank acknowledge that they've promised a lot and that to some extent they're figuring things out as they go. And we knew this, right? They announced this.

1:21:16

It was like, you know, Elon was pointing it out.

1:21:17

He was saying, "Where's the money coming from?"

1:21:19

Yeah, it's very, you know, pulling $500 billion together is not uh not the easiest thing in the world.

1:21:26

Um, so they dis they dismiss critics.

1:21:28

The loudest being Elon Musk who says the full scope will never be realized because they don't have the money. We don't.

1:21:34

And Masa says uh founder of SoftBank uh the infamous Japanese piggy bank is this article in the economist says we don't need $500 billion in one day. We'll go step by step.

1:21:46

Do you think more founders should be pitching stuff like Stargates?

1:21:51

Like yeah like I I need to realize my full vision. I need 500 billion.

1:21:54

I don't need it right now.

1:21:55

So like if you're if you're going to come at me for that. That's ridiculous.

1:21:59

It's like I didn't say I needed it right now, but over the course of this company, we will consume 500 billion capital. Yeah.

1:22:04

I mean, a lot of the nuclear companies might, you know, who knows? We'll produce trillions.

1:22:09

So, OpenAI CEO Sam Alman knew they'd need construction crews and lots of computers. Yep.

1:22:14

We're going to need computers.

1:22:16

Hey folks, we're going to need lots of computers.

1:22:18

We're going to need computers.

1:22:20

But data centers also have extensive electricity and water needs to keep the machines running at a safe temperature, often requiring infrastructure upgrades that can take years.

1:22:29

For Abalene, which will have mighty capacity of 1.

1:22:31

2 gawatt, the team built its own gas power plant to get the place running more quickly. Yeah.

1:22:37

that the the we talked a little bit about Augustus, but there's this whole meme about like every time you search JBT it like pour empties out an entire lake and like we're running out of water and like I I mean I'm sure it uses some water, but I have no idea exactly how that's calculated and that feels like something that's Yeah.

1:22:54

The question is what percentage evaporates during cooling process? Yeah.

1:23:00

And is it like a lot of a lot of cooling systems are pretty closed loop because you're just piping you're piping the water through through tubes and then the water goes onto the chip.

1:23:12

The chip heats up, the water goes the water gets hot and then you pipe the water out.

1:23:15

That's like a a water cooled GPU rig on like a gaming PC.

1:23:18

And like yeah, you need to change out the water every once in a while, but like not constantly.

1:23:24

And so it's not like I I I I do wonder how what the actual impact scale is here.

1:23:29

And I bet there's a lot of people with like the incentive to say don't use any water or way too much water.

1:23:35

Wonder I wonder if you could rip out the the parts of an air cooled Porsche and potentially use those. Yeah. Air coo.

1:23:42

I want an aircooled and then Elon could take air cooled Porsches and put uh engine swap Tesla motors into that.

1:23:50

I I I think we wanted supercharged AI factory.

1:23:53

Just pipe the pipe the air back in. Yeah. To accelerate it.

1:23:55

uh take the exhaust fans and reroute them right onto the chip. Yeah.

1:24:00

So anyways, Sam knows he's going to need a lot of computers.

1:24:04

Uh but they also have extensive electricity and water needs.

1:24:09

So Abene is going to have 1. 2 gawatt.

1:24:12

The team built its own gas power plant to get the place running more quickly.

1:24:16

One of the things that really surprised me as we were starting to dig into this, Alman says, was just how many things feed into the mainline.

1:24:21

The main line is still being assembled at the Abalene site in March.

1:24:25

Some rooms are in the process of being wired and are off limits due to the risk of electrocution.

1:24:31

In the future, tubes will pump cooling liquid around servers which will be loaded with Nvidia graphics or sorry GPUs and to the physical GPU actually stands for graphics processing unit as they thoughtfully noted in this piece. It's good to note.

1:24:44

It's good to note uh if you've been living under a cluster a GPU uh they they also contextualize it by telling you what a GPU here is.

1:24:57

It's not only a graphics processing unit.

1:24:59

You can think of it as the physical brain of AI something there. Something there. PB PBAs.

1:25:07

This is this is this is ripped straight from a soft bank deck. Yeah.

1:25:11

It's like they're talking to the same audience. Retail. Yeah.

1:25:12

For now, those tubes dangle from the ceilings beside half-lit hallways.

1:25:18

We're trying to deliver on the fastest schedule that a 100 megawatt or greater data center has ever been built.

1:25:25

Lock Miller says from the front seat of a buggy on a ride through the construction site.

1:25:28

In this moment, speed matters a lot.

1:25:30

The big reveal for Stargate was at the White House on day two of Trump's second term.

1:25:36

I didn't I forgot it was so early.

1:25:38

Altman Sone and Oracle chairman Larry Ellison stood behind.

1:25:40

This was so funny too, remember?

1:25:42

because it was it was Elon and Sam like both around the White House on the same day in the midst of this battle.

1:25:50

Dude, a little bit of an awkward Have you have you seen this this picture of Masa getting height mogged so bad by Larry? It's brutal. Really, John?

1:26:01

You're going to bring up height mogging?

1:26:03

It's a sensitive topic on the show.

1:26:03

It's a sensitive topic, but we but I mean we we should pull we should pull it up.

1:26:07

We got to call it out anyway.

1:26:09

um lavishing praise and crediting him with help making the pos the project possible.

1:26:13

The whole production came together in a matter of days according to a person familiar with the events planning who asked not to be identified.

1:26:21

Um, Stargate's real origin, according to Peter, the vice president of infrastructure and strategy at oper and operations at OpenAI, uh, Hosle Hosche, I'm not sure how to pronounce the last name, uh, goes back to a research paper published at the beginning of the decade written by a group at OpenAI that included Daario Amade, who'd go on to start one of the company's main rivals, Enthropic PBC.

1:26:42

Um, the paper described so-called scaling laws.

1:26:46

This is GPT2 which we've talked about a bunch uh which assumed that the more capable AI requires ever more data and computing resources.

1:26:54

What's interesting is like we're talking about increasing thems uh but like this has to be synthetic data at this point because we've kind of already hit the data wall I believe or are we because I I I don't know that there's ever anyone expanding it.

1:27:05

Yeah, but I I don't know that anyone's really mapped out like 10x uh Yeah, check out that photo.

1:27:14

Um, rough podium situation.

1:27:14

Trump's also pretty big, too.

1:27:18

Larry's pretty big, but uh, yeah, Masa, well, tough presentation. Still a size lord.

1:27:26

Yeah, makes up for it with the checks he rips.

1:27:28

Yeah, OpenAI uh, as many know, already had a contract designating Microsoft, its primary financial backer, as its exclusive cloud provider, but Altman decided he'd eventually need other options.

1:27:40

Oracle was looking for partners for the future complex in Abalene and was talking with Musk before Altman came in says a person familiar with the discussion.

1:27:47

Uh Musk obviously chose Memphis uh for XAI so went a different direction. Oh, this is cool.

1:27:56

Alman picked the name Stargate because of one of OpenAI's early data center designs resembled the ringshaped portal that could open a wormhole in the 1990s science fiction movie of the same name. Also a TV show.

1:28:07

Uh you ever watch Stargate? Absolutely not. Next question.

1:28:12

After half a year of Altman selling the idea to prospective backers, Stargate really came into focus in mid 2024.

1:28:17

And he says he found he found himself in a lot of Zoom meetings.

1:28:22

Stargate also a CIA project around mind reading. Oh, I don't know.

1:28:27

I'm not familiar with that one. Look it up.

1:28:28

Chad can The hardest part was figuring out what shape of the what the shape of the deal would be.

1:28:34

Would it be part of OpenAI?

1:28:34

Should it be a separate entity?

1:28:36

You know, they love separate entities over at OpenAI.

1:28:37

Oh, we know where they landed on this one.

1:28:41

Should we just try to get one company to build it all for us?

1:28:43

That's the Stargate project was a classified United States Army program.

1:28:47

Wasn't affiliated with the CIA, but it was initiated in 1977 by the DIA and SRRI International to explore the potential of psychic phenomena in intelligence gathering and military applications. How'd it go?

1:29:00

Did they were they successful? I mean, seems like it.

1:29:05

I'm sure I'm sure they I'm sure they cooked.

1:29:06

So, OpenAI is responsible for the operation of the business and will be its main customer.

1:29:10

SoftBank handles the financial side, including raising additional funds for Abene.

1:29:14

Oracle will lease the data center and fill it with servers.

1:29:17

Crusoe declined to comment on who's else is going in.

1:29:19

Uh, Openhand and Soft Bank will each put in 19 billion to start.

1:29:25

Oracle and another equity partner, Abu Dhabi based investment firm, MGX, are on the hook for 7 billion a piece according to a person familiar with the matter. So, um, yeah.

1:29:33

I mean, 19 and 19 plus seven and seven, you're getting up into the 40s, 50s, 60s, billions, not bad.

1:29:41

It's not 500, but they're getting there. They're within one.

1:29:45

That's what they got to do. Striking distance.

1:29:47

Each project will be financed individually with a mix of equity and debt uh with uh data center firm Primary Digital Infrastructure is helping orchestrate the financing.

1:29:55

JP Morgan Chase is uh doing some loans. Yeah.

1:30:01

So if you have money over at Chase, you're I would say at this point entitled to say that you're a financeier of Abene of the of the project in Abalene. That's great.

1:30:10

For future sites, SoftBank has held talks with dozens of lenders and alternative asset managers.

1:30:17

Although it's yet to secure a deal and some of those conversations have slowed due to global market volatility, but now we're kind of flat.

1:30:23

Maybe the markets open back up and things move on.

1:30:25

That's a structured financing that SoftBank has done many times in the past.

1:30:29

So we are very confident and yeah I mean SoftBank took ARM private and so they're clearly op they're clearly capable of operating at this scale. It's not not crazy.

1:30:35

Uh not everyone is so optimistic about SoftBank's ability to execute.

1:30:38

The last time SO did anything close to this scale was the vision fund which depended on a $45 billion contribution from Saudi Arabia's largest sovereign wealth fund.

1:30:48

Imagine clipping two and 20 on that.

1:30:50

Soft banks fund racked up losses most disastrally most disastrously on the soft real estate startup.

1:30:55

Sometimes I get crazily over excited and I make a mistake like we work.

1:31:01

But when you have the conviction and the passion's mistake was not backing what's the new company that no no no the new real estate company uh by Newman.

1:31:11

Oh I don't know did flow flow he should have doubled down but when you have the conviction in the past and you actually learn from those mistakes and the scars and that will make you stronger.

1:31:22

I've made so many more mistakes that I think I'm a little stronger than in the past.

1:31:27

He's been very wrong, but he's also been very right.

1:31:29

And he's been right more than he's wrong.

1:31:31

And so he's still in the game.

1:31:32

And as long as SoftBank is cooking as a company, he can keep ripping checks. Yeah.

1:31:36

I mean, it's his calculus is I believe can be done on a napkin.

1:31:42

And this is just my guess. Huge napkin.

1:31:45

He has a very large napkin and it's basically if I buy 10% if I buy 10% of open AAI and it's a trillion dollar company someday I'm going to do very well and I'm going to lever up to do that.

1:31:58

So my actual return on equity is going to be fantastic and he's going to be laughing and this is I mean this is like this is like it's not as crazy some of the best books on Masa one of the best books on Masa is called the gambling man.

1:32:12

That's the context you need.

1:32:13

Yeah, we did a whole deep dive on it.

1:32:15

Complicating matters are Trump's tariffs.

1:32:18

Cost calculations swing every time the president changes his mind, but the trend line is expected to go up on uh is expected to go up on aluminum, steel, and other construction supplies, not to mention Nvidia GPUs, which are made mainly by Taiwan Semiconductor Manufacturing Co. TSMC.

1:32:31

Uh S says he expects no significant impact from the trade policies. And I think he's right.

1:32:36

I think this is what Trump wants.

1:32:38

I think all of these uh will get will get carve routes and they've already been negotiated down.

1:32:42

So I I'm not too worried about that.

1:32:44

Um Trump I I think the bigger question here is just like like what is the return to scale because we've seen the scaling laws kind of uh maybe potentially drop in terms of pre-training wall and whatnot.

1:32:55

And so the question is just like what does a 10x larger facility actually get you?

1:32:59

it better be worth it because we know the we know the economics of GPT4 training run were great.

1:33:08

Like they they printed tons and tons of $20 a month subscriptions off of that and I think the total cost was like $100 million.

1:33:16

Now you're up at 10 billion.

1:33:16

Can you get everyone on $2,000 a month subscription? If you can, no big deal. Well, yeah.

1:33:20

And a lot of I mean, won't a huge amount of Stargate resources just go towards servicing Yeah. models potentially.

1:33:28

I mean I I feel like a lot of the narrative is like training but you know obviously obviously it will be both.

1:33:35

Um Alibaba chairman Joe Sai has raised the possibility that we're in the middle of a data center construction bubble uh and uh obviously Amazon Microsoft have pulled back on some data center plans.

1:33:51

Alman is undeterred still seeing compute capacity as precious resource.

1:33:54

We need more compute and more capital.

1:33:55

He says, "We want to have we want to have access to a lot more of the machinery to make AI and run AI."

1:34:01

And I mean, yeah, like the GPUs do go on fire and there are limited resources still even with the models.

1:34:10

I mean, obviously there's a ton that you can do on compress.

1:34:11

The fact that you're getting rate limited by Google as a paid customer, it's insane. Yeah, it's insane.

1:34:16

Like it is the biggest bull case for build more Yeah. data centers.

1:34:20

But I I I honestly put that less on on on their cloud offering than just some PM was like, "Oh, no one would want to generate more than three videos in one day or whatever." Totally.

1:34:33

Or like, "Let's just roll this out slowly."

1:34:34

Um, Alman is undeterred, still seeing computing capacity.

1:34:39

Uh, the Abene Skyline is made up largely of buildings constructed in the 1920s during the Texas oil boom.

1:34:44

Among the city's major employers is a nearby Air Force base and cheese production plant that opened in 2022.

1:34:52

Uh meanwhile, activity at the Stargate site spanning 900 acres larger than New York City's Central Park moves at the pace of a boom town.

1:34:59

And I was I was talking to some real estate folks and I was saying like you should probably just go to Abalene and start building like Walmarts and gas stations and everything because like it really will be a boom town.

1:35:10

They're going to be building this for a long time.

1:35:11

Yes, these are going to be somewhat human resource light, but there still will be a ton of people working there, ton of people getting paid a lot.

1:35:20

They'll want bars and restaurants and like the local economy should like it would be impossible for it not to flourish.

1:35:26

Like the economy booms when you put in like a soccer stadium.

1:35:30

Like you're talking about Yeah.

1:35:32

I wonder if there's I wonder if there's been any I wonder if there's been any people going I mean I'm sure the Stargate project in general has just been buying up surrounding land to just sort of mitigate potential risks or future issues.

1:35:45

But if you own like an acre, even just a single acre that's not connected to roads at all within this area, you could probably end up selling it for Yeah, that's smart. You know, millions.

1:35:57

You got to just hold on and make them build the data center around you.

1:36:00

So you just do you need a helicopter to get helicopter and you just look every you have this quaint little house on the prairie and then it's own little microclimate because like the heat coming off.

1:36:09

Have you seen that photo in Virginia where basically there was like this nice little house and then AWS and Amazon just built up these massive and so you just see this nice house and you pan up and it's just as far as the I can see data centers.

1:36:24

It's like beautiful rough.

1:36:24

The team behind the Abene data center conceived under the code name Project Ludicrous.

1:36:32

Nice, nice code name, first put shovels on the ground in June of 2024.

1:36:34

Crusoe decided to fly in workers from all over the country because of the city's uh because the city of 129,000 people didn't have a labor force large enough to support it.

1:36:44

Lock Miller says uh its project 129,000 is way more than I had originally kind of imagined for this.

1:36:52

probably got some good high school football.

1:36:53

That feels more like a city than people people like to position Abene as like oh this quaint little town.

1:37:00

Um it's projected to take slightly more than two years for the entire project to go from start to finish.

1:37:06

Although parts of the data center are slated to go online sooner.

1:37:10

Towns like Abalene are usually on the losing end of technological change which tends to create most jobs in the largest cities and data centers traditionally don't employ that many people.

1:37:19

Crusoe has promised officials that the Abene site will deliver 357 full-time jobs once construction has finished.

1:37:26

That's not that much actually. That's pretty small.

1:37:29

I thought it'd be higher for a $500 billion project.

1:37:30

Well, I guess it's 10 billion in this version or what do they 10 or 50?

1:37:34

But yeah, that is crazy low numbers, but the AI agents hate to be contributing to this.

1:37:41

I backed a company that makes robots for data centers. Reducing it from 350.

1:37:49

We're going to get that number down to five. Yeah, that's the goal. Yeah.

1:37:53

I mean, it's basically the idea that like data centers are flat and there's like routine maintenance that needs to be done.

1:38:00

It's reasonable and it's like very much a controlled environment and just driving a little Wall-E style robot around makes a lot of sense.

1:38:09

It's like we're talking about like how much of a boom town it is.

1:38:11

It's like 375 jobs for now on a base of 129,000 existing people.

1:38:17

I I feel like there will be more knock-on effects though.

1:38:22

Anyway, the What about the There's Chipotle opportunities. Yeah.

1:38:26

Those 357 people are going to be eating a lot of double steak guac burrito bowls. Yeah.

1:38:31

So, the Abene government is hopeful that it is the start of something bigger.

1:38:37

The city and county agreed to extend significant tax breaks to get the deal done.

1:38:40

And local officials saw an influx of requests in the days after Stargate was announced from housing developers, infrastructure, energy infrastructure vendors, wouldbe investors, and other municipal governments desperate to bring jobs to their town.

1:38:51

Um, it will impact the rest of the economy.

1:38:53

Our restaurants, our home builders, our Chipotles with many new people coming in and taking these jobs.

1:38:58

The city was able to accommodate Stargate's water needs, which are small compared to a typical server farm due to a novel cooling method that recycles most of the liquid.

1:39:09

Let's hear it for the fake news that Chad GBT is destroying the ocean. It's not true.

1:39:14

Like, they figured out how to recycle the water, of course, cuz water's expensive and why would you waste it?

1:39:20

The trade-off is that the data center it's not the water is not being massively contaminated like in oil production fracking and so they need uh a ton of electricity data centers at this scale are unprecedented.

1:39:32

Uh the this was caused for concern where the memory of 20 2021 Texas blackouts which led to the death of six people in the county uh is still fresh in people's minds.

1:39:43

Much of the blame for those outages belongs to Texas's deregulated energy grid, but the flexibility that comes with the lack of regulation was part of the allure for Stargate.

1:39:53

Crusoe went out and bought gas turbines and oversaw the construction of a natural gas plant on site, which will supplement the what the data center can get from local utilities.

1:40:04

Yeah, I talked to an investor who uh was uh who knows Chase and was like he's just he has an uncanny ability to find energy which is like I don't know something I wouldn't even think about in terms of like what the skills of a founder are necessary but he's like extremely good at finding like opportunities and pockets of energy that are being underutilized.

1:40:23

So I'm technically millennial but I'm was fairly close to being Gen Z by a few days. Yep.

1:40:30

So is that like rising energy?

1:40:32

Like I'm just trying to put it He's an He's an energy. No, it's like Rizzler.

1:40:37

Is that what you're talking?

1:40:38

It's like it's like there I think there are a lot of there are a lot of companies out there that think kind of along the lines of like a single energy source.

1:40:44

So there is a lot of of natural gas here.

1:40:48

Chase is very good at putting together different opportunities.

1:40:52

There's a existing plant, we can build a new plant, there's wind, solar, like very energy agnostic, but how do you piece all those together to create the act to meet the actual demand on an ongoing basis?

1:41:03

Uh, Stargates to Stargate critics, the idea that the project's backers could maintain this base and scale across the country is reason to think they're living in a fantasy.

1:41:12

Amade, co-creator of the scaling laws, publicly questioned Stargate seriousness, calling it chaotic.

1:41:16

Musk, breaking with Trump, called it fake. Altman's retort.

1:41:21

He says all sorts of crazy stuff.

1:41:23

Salesforce uh CEO Mark Beni off said in a post on X in April that Stargate signaled an end to the honeymoon between OpenAI and Microsoft.

1:41:32

The project does reflect a shift in the relationship between the two companies.

1:41:35

They revised their cloud contract in January to allow OpenAI to use other vendors such as Oracle as long as Microsoft has still has right of first refusal.

1:41:43

Microsoft doesn't intend to invest in Stargate but was listed by OpenAI as a technology partner in the project.

1:41:48

And this was interesting because uh they Can I just put your logo on the deck, please?

1:41:53

Can I please just put your logo on the deck?

1:41:55

Yeah, there's there's this amazing Microsoft is providing Exchange.

1:42:00

There's this amazing There's this amazing line uh from this guy Aaron, buddy of mine.

1:42:05

We'll be using Microsoft Excel on site.

1:42:07

So, they're a technology partner. Yeah, that's it.

1:42:12

Um quote, "I'm good for my 80 billion."

1:42:14

He's like, "I'm good for my 80 billion generically capex. It'll happen somewhere.

1:42:18

It's not necessarily an investment starting.

1:42:20

I'm going to butcher the exact quote, but Aaron Frank has an iconic post at one point, or maybe he just said this to me in person.

1:42:27

He just said, "Some people just want to see my name in their some people just want my name in their deck."

1:42:31

Because there was like this fintech boom and people would just come to him and be like, "Please, can I give you 50 bips to just like put your name?"

1:42:38

Because like that like at some point, what's his name?

1:42:42

He he created the card that became the Apple. Oh, okay.

1:42:45

What what was it called again?

1:42:48

You don't even see that the Apple push. No, Apple Card. Remember Apple Card?

1:42:54

I think it was called the Apple Card, but his his company like became that. Yeah, Apple Card.

1:42:57

You don't hear much about the Apple Card.

1:42:59

They had a deal with Goldman and then they kind of like fell off and maybe they're not doing it anymore. Did people get those? I never had one. Was it good?

1:43:05

Did it have like cash back on stuff? It was pretty good. Okay.

1:43:09

Anyway, unlimited daily cash back.

1:43:11

I'm not really a credit card points maxer or into credit.

1:43:14

been big into that world.

1:43:17

Never too uh management consulting coded.

1:43:19

It's like every uh every guy I know who works at like Mackenzie or Bane or BCG is like a spreadsheet points.

1:43:26

I'm just like I don't care.

1:43:28

Just tell me which card I'll get and I'll use that for a decade and they'll they'll squeeze me for annual fee probably and I'm probably like way underwater on it because I'm not points maxing. Yep.

1:43:37

Anyway, uh if things go as planned, Abene could serve as a model for the next of Stargate's developments.

1:43:43

The day after Altman first visited the Abene site in early May, he testified before Congress asking to for help to fasttrack data center permitting on future locations.

1:43:50

The f the first site was incredible.

1:43:52

We need a lot more of that.

1:43:54

Crusoe is evaluating another potential spot in Amarillo, Texas, which it hopes could be part of Stargate.

1:44:00

According to a person with knowledge of deliberations who asked not to be named, OpenAI has said is also looking in several states including Oregon, Pennsylvania, and Wisconsin.

1:44:07

And while Stargate might rent data centers already built being built elsewhere, uh meanwhile, OpenAI and Oracle are taking the spirit of Stargate International with an AI data center under development in Abu Dhabi, which will not be part of Stargate LLC, but will count OpenAI as a customer. It's uh simple.

1:44:24

Uh if all that wasn't enough, Stargate could even expand to include semiconductor production, according to Altman, perhaps as soon as next year.

1:44:33

I think of Stargate as the AI factory.

1:44:34

If we don't have to be involved here, maybe we wouldn't.

1:44:39

But if we could just ma magically spin up all the compute that we needed in the sky.

1:44:42

Um, but at this kind of scale, he can't do that.

1:44:45

And so there's another interesting angle here, which is uh the idea of building an AI data center in space.

1:44:52

Uh, we're having a couple folks to come on the show maybe next week to talk about that.

1:44:56

There's a company that's thinking about doing it.

1:44:59

There's some investors that are looking into it. It's sounds silly.

1:45:01

Sounds extremely buzzworthy and sci-fi.

1:45:03

It's like hard tech space and AI and data centers. Um, but think about it.

1:45:08

You get you get energy from the sun from solar panels basically for free.

1:45:14

And then you also get cooling because you're in the vacuum of space and it's cold for free.

1:45:20

And so you don't need probably water cooling and you don't need as much energy and so the economics might work out potentially in GP pretty yeah you don't need land and so you've cut out a bunch of costs and then added Starship or or Falcon 9 launch costs which are significant but maybe small.

1:45:36

I cannot wait for the day where a satellite is gibbling for me.

1:45:43

Just knowing that I can create the most silly picture in my mind and then send it up to space to get made and then have it sent down.

1:45:53

Just send it back whenever you get around to it. Y could be tomorrow.

1:45:57

And then it says, "You've run out of requests. We don't do anymore."

1:46:01

When's the last time you saw Gibli? It's been a while.

1:46:04

I have been using um I've been using images in chat GP chat GPT a lot but on a a lot more like practical sense just trying to visualize different things. I like the library.

1:46:18

I told you I've been using I've been using Chad GPT image uh images to uh get my three-year-old to eat dinner.

1:46:29

You said that bite by bite.

1:46:31

I will generate a new if you take another bite of chicken uh and chew it well, I'll generate another picture of me as a dinosaur.

1:46:38

My new my new go-to Gibli instead of asking turn this into a studio Gibli.

1:46:44

I just say uh turn these turn the image into bodybuilders.

1:46:48

And I've been having a lot of luck with that.

1:46:51

I I I sent one to the chat.

1:46:51

We'll try and pull it up.

1:46:53

But uh get this pulled up. It's it's great.

1:46:55

Oh, but that's that's the original picture.

1:47:01

Yeah, this is so we'll pull up the original picture first of the Open AI team. There they are. Okay.

1:47:04

And then can you can now can you make them look like bodybuilders bodybuilders? Yeah, for sure. Um yeah. Yeah, I enjoy that.

1:47:12

Um also just making Yeah, just a lot of a lot of Cadillac Escalades with uh GT3 liveries and uh and a lot of uh TBPN merch ideas, a lot of stringers, a lot of suits with with ads on it.

1:47:27

Um, speaking of which, we should do some ads. Uh, go to vanta. com.

1:47:32

Automate compliance, manage risk, prove trust continuously.

1:47:35

Vantage trust management platform.

1:47:37

Takes the manual work out of your security and compliance process.

1:47:39

If you're doing anything important, you probably should be on Vanta for sure. You heard it here first. Go to vant. com first.

1:47:47

Should also get on numeral sales tax on autopilot.

1:47:49

Spend less than five minutes per month on sales tax compliance. Benchmark series A. Benchmark series A.

1:47:54

Anyway, uh, do we have time to do Wait, well, we only have 15 minutes until our next guest. We do a little timeline.

1:48:02

We should do a little timeline. Be good.

1:48:03

We got Austin Allred coming in.

1:48:05

We're going to ask him about AI coding in the era of agents. Uh, we can cover Daario. Maybe tomorrow.

1:48:10

I also want to go through Oh, we didn't get a chance to dig through um Craig Federigi and the Apple AI uh piece in Bloomberg, but there are some really hilarious uh points in here.

1:48:27

One is that uh apparently when Steve Jobs was trying to acquire Siri, he called the CEO and said, "I'd like to buy the company.

1:48:36

I think this is the future of computing.

1:48:38

I think voice interfaces, artificial intelligence, this is what Apple needs.

1:48:43

We want to buy your company. Bring them in house.

1:48:44

Uh the CEO of Siri said, "No, I don't want to be part of Apple. I'm good."

1:48:49

Uh and Steve Jobs proceeded to call him every single day for 24 days straight before he got the deal done. Insane. Absolutely insane.

1:48:56

You don't hear about that a lot.

1:48:58

Uh yeah, you you have you get this idea in your head of of somebody that is a demigod figure and you think that they don't have to be cringe to get things done. Yep.

1:49:10

And turns out sometimes you got to cringe Max. Yeah.

1:49:16

And and do whatever it takes and sometimes that's calling somebody every single day.

1:49:19

It must just be so funny to pick up the phone. Imagine you're like 24.

1:49:24

Imagine you're like Yeah.

1:49:24

walking your dog or you know it's like Steve Jobs calling again.

1:49:29

Oh, he's leaving a voicemail. Oh, hey Steve. How you doing? Still not interested.

1:49:34

Still building my company. Still grinding.

1:49:37

Steve Jobs would get sort of total privileges to call people bud.

1:49:45

Yeah, totally one of the few when you're at the top of one of the few that earned it.

1:49:49

When you're the greatest entrepreneur of all time.

1:49:50

So if you're not Steve Jobs, don't use remove that from your vocabulary.

1:49:53

Uh, let's go to some timeline.

1:49:56

Zack Kukov's talking about, "Oh my god, we're really going to speedrun 2008 with collateralized losses have widened as more consumers fail to repay loans.

1:50:09

Unfortunate learning, you know, basically probably speedrunning uh risk management that credit card companies have done for a really long time.

1:50:19

I wonder what's actually going on here because the whole like yes, you can buy now pay later on your burrito.

1:50:25

That's silly, but that's probably not the vast majority of their underwriting, right?

1:50:29

Like that's the meme, but that's not actually what's driving uh buy now pay later activity.

1:50:33

And I wonder if in the leadup to IPO, they they loosened underwriting restrictions to try and ramp volume and because it seems like CLA, have they gotten out? No. Right.

1:50:46

They they they're not public yet, right?

1:50:48

they have not successfully IPOed.

1:50:48

And so I imagine that in the leadup to the IPO, they're trying to do a bunch of things.

1:50:54

So there's that whole that whole news cycle around, oh, we're not going to hire anyone.

1:50:57

We're going to use AI for everything.

1:50:58

Obviously, that's sending a signal to the market that, hey, we're going to be an ultraefficient business.

1:51:03

The margins are going to dramatically improve.

1:51:05

We are taking AI super seriously at this company.

1:51:07

And so, uh, you know, our opex should be really low over the long term.

1:51:12

you can underwrite the the IPO against a really low uh SGNA cost, right? Yeah.

1:51:18

Um and then also on the revenue side, they're probably saying they're probably trying to do everything they can.

1:51:24

Maybe they went too far and now they're uh generating losses, which is obviously not Yeah.

1:51:28

So, the issue is they doubled losses.

1:51:30

So, their net loss for the first three months of 2025 was around 100 million.

1:51:34

A year ago, for the same period, it was 47 million.

1:51:39

Is this from like the S1 or something?

1:51:42

Like why is this data public?

1:51:46

Uh I think they're they're trying to get out and so I imagine they need to do some type of reporting.

1:51:55

Revenue increased 13% year-over-year to 700 million for the quarter.

1:52:00

And again, this is a company that in 2021 was valued at $45 billion.

1:52:05

Most recent private mark was around 15.

1:52:09

So unclear what they'll do.

1:52:09

They put their IPO plans on hold last month along with StubHub.

1:52:15

And again, they had been on this marketing tear, right?

1:52:19

Remember the CEO came out and he was like, "We're using AI for everything.

1:52:23

We're we're letting everyone go."

1:52:25

And then he says, "Actually, no.

1:52:26

We're we're not doing that.

1:52:28

It's not good enough yet."

1:52:31

I don't like when the IPL windows closed. I like an open window. Yeah.

1:52:33

I like companies going out every day.

1:52:36

I like you punched a hole in the wall at the gym when you heard I remember when the window closed when it closed. Yeah.

1:52:45

I was like, "John, why is your hand two feet into or sorry, why why is your fist two feet into the wall?

1:52:50

John, you were trying to get it out.

1:52:52

Too frustrated, pissed off.

1:52:56

Well, if you're trying to track all these IPOs, trying to get in on the action, head over to public. com.

1:53:01

Investing for those who take it seriously.

1:53:03

They got multi-asset investing, industryleading yields, and they're trusted by millions.

1:53:06

Also, Jordy, how do you sleep last night?

1:53:08

I put up some good numbers.

1:53:10

I'm climbing up the ranks.

1:53:12

I'm getting back in the game.

1:53:14

I might have beaten you back days. Let's see. How'd you do? What's your number? What's mine? No, you go first. Okay. Well, one second.

1:53:27

Um, get a Pod Five Ultra.

1:53:27

They got a fiveyear warranty, a 30- night risk-free trial. Got an 80.

1:53:31

John, I'm sure you beat me. Oh, I got a 94. Smoked. Smoked. Get out of here. Get out of here. Retry. 96. Actually, 96. 96. Weep. Boom. 96. Good to see you, John. I'm proud of you. Yeah.

1:53:46

Still not good on the consistency.

1:53:48

Going to bed at all random times.

1:53:50

Waking up pretty consistently. 5:45. Uh anyway. All right. Going on.

1:53:55

Uh we kind of covered this already.

1:53:57

Uh this from a lion does not concern himself with sock 2 compliance.

1:54:02

That's because the lion uses Vanta, right? Exactly.

1:54:04

You don't need to be concerned.

1:54:07

Maybe we need to promote this.

1:54:09

Yeah, this is an ad for this is an ad for sure.

1:54:11

Andy, you got to be disclosing.

1:54:13

Yeah, you have to do hashtag ad or maybe he just loves Vanta so much. It's possible. It could just be UGC. Yeah, could be.

1:54:19

But anyway, obviously don't concern yourself with sock 2.

1:54:21

Be a lion and get on Vanta.

1:54:23

Uh Art over at Brex announced a partnership with Zip, a big procurement platform.

1:54:30

Apparently, Zip is taking a step back from the corporate card market and so this partnership uh allows them to use Oh, interesting.

1:54:39

Okay, that that makes a lot more sense.

1:54:40

So, Zip had a corporate card and now they're just partnering with with Brex and B had a procurement product. Yeah. Which they rolled back.

1:54:48

And that kind of happened too with with Stripe had a corporate card for a little bit and then stopped that, right? Isn't that what Yeah.

1:54:52

Everybody was like, "Wait, corporate cards, good business.

1:54:56

We should do corporate cards."

1:54:56

Turns out hard to hard to do. Yeah.

1:54:59

Well, there's always a question about like, "Where does the corporate card sit?"

1:55:02

Because Stripe does not sit in the same place as like Ramp in terms of like the CFO suite.

1:55:08

Like it's a it's more of like almost a developer tool.

1:55:10

It's less their initial wedge, but Yeah. Yeah.

1:55:12

But it's it's less about like an accounting product.

1:55:16

It's more about like the like the the payments and engineering product.

1:55:21

Uh although obviously it's very tied but I don't think Stripe's ever gone into travel or you know bill pay like they they haven't focused on that as much.

1:55:30

Anyway, uh Art says this is two YC companies with cultures of shipping fast and collaboration creating something truly special.

1:55:37

We love to see YC companies.

1:55:39

YC companies. Obviously the most f the recent collaboration between deal and rippling also YC companies hasn't that yeah a little tumultuous necessarily call it a collaboration well they were partnered in some ways because they were both employee yeah

1:55:56

they were the same employee anyway uh back in 2021 uh art approached ZIP CEO CEO because I saw Zip doing something remarkable in the enterprise segment a market Brex was determined to At the time, uh, the CEO of ZIP pushed back saying, "Rex was too focused on startups. What would he get from this?"

1:56:13

What would he get from this?"

1:56:15

Fast forward to March 2024.

1:56:15

Brex has built their base of enterprises, and many were demanding our products to work seamlessly together.

1:56:21

Both companies had also retrenched to focus on core strength.

1:56:24

Zip stepped back from card.

1:56:24

We step back from enterprise procurement.

1:56:28

This is a classic competitors turned co-creator story.

1:56:30

ZIP brings world-class procurement orchestration.

1:56:33

Brex brings global card capabilities and spend management across 50 countries.

1:56:37

So, congratulations to Art and Bra.

1:56:39

Um, Art, fellow uh LA guy.

1:56:42

And I have some advice for Bra, why don't you rent a billboard to promote this on Adquick?

1:56:51

Yeah, get some Brax actually has a ton of billboards, right?

1:56:56

They I mean, they ran They invented the billboard in many ways.

1:56:59

They did invent the billboard.

1:56:59

They had billboard economy on their back in 2020, 2019, right?

1:57:04

It was like it was hard to get a billboard because they were on all of them. So head over to adqu. com.

1:57:08

Art uh out of home advertising made easy and measurable.

1:57:12

Say goodbye to the headaches of out of home advertising.

1:57:13

Only AdQuick combines technology out of home expertise and data to enable efficient seamless ad buying across the globe. Double kill. Great.

1:57:23

Uh great transition, John.

1:57:23

Uh we got a post here from TJ Parker who will be on the show Thursday again. He's coming back on.

1:57:30

and he says, "One thing I'll never understand about healthcare startups is the instinct to take a strategic route rather than an administrative tactical approach.

1:57:36

You don't need a top-to- meeting to get in network with a payer.

1:57:40

Just fill out the paperwork.

1:57:42

You don't need a strategic relationship with Quest LabCore to integrate scheduling or get competitive pricing.

1:57:48

Just work with an existing vendor.

1:57:50

If things go well, you can elevate to strategic relationship over time, but you'll never get off the starting line with the strategic approach."

1:57:57

So, uh, Will has a has a funny reply here.

1:58:02

It's very funny when people tell me they're struggling with getting in to in touch with payers.

1:58:05

It's like, have you tried going in the front door?

1:58:11

Timeless advice, timeless wisdom.

1:58:11

I wonder what's going on here.

1:58:13

Is this just because there's a lot of like tech people who think I'm going to build tech for healthcare and then they're just completely unfamiliar with think it's all about partnerships versus just or they're just unfamiliar with this market structure because TJ's been I mean he was a pharmacist before right like he really understands the industry and has continued to understand the industry even post exit.

1:58:32

Um, whereas I think a lot of people are like, I build software.

1:58:36

I know that healthcare is a problem, so I'm going to build healthcare software, but they don't understand like the structure of the market at all or how the deals get done.

1:58:45

And so, they got to learn stuff from from TJ's posts, I guess.

1:58:47

Anyway, we got a post here from Andrew Reed.

1:58:50

Uh, in the news, Authentic Brands buys Dockers from Levi Strauss.

1:58:56

A little mixup in the pants market.

1:58:59

[Music] Little switch up.

1:59:01

And Andrew says, "Docker, the containers, uh, overlapping with Docker's the pants.

1:59:06

Both making hundreds of millions, selling to system admins." Um, yeah.

1:59:11

You know, Andrew is a great uh he's a great investor, but he puts his pants on one leg at a time like anyone else. Yep.

1:59:19

Uh, let's get him some Dockers. Let's see. Get him some Dockers.

1:59:22

I somehow think he's We should get him too stylish for that. Yeah.

1:59:24

I can't see him wearing Dockers, but we should get him a TBPN tie. Yeah, tie would be good.

1:59:31

working on with ads on last time with ads.

1:59:33

Yeah, last time he came on the show, he was he was wearing a actually really nice Figma sweatshirt, but didn't have a collar, so you know, docked some points from that. Yep. Docked. Docked.

1:59:44

Uh well, you know, we should also tell you about wander. Find your happy place. Find your happy place.

1:59:54

Book a wanderer with inspiring views, hotel grade amenities, dreamy beds, top tier cleaning, and 24/7 concier services.

1:59:58

Vacation home but better, folks.

2:00:00

Um John Andrew founder Wander was teasing how excited he is about some big news and I couldn't possibly guess what the news is but we're going to have him on the show next week show says OpenAI tried to partner with Google to use its search API to part to power real-time results in chat's search offering.

2:00:17

Google said no, citing too many complexities.

2:00:19

And this is from uh Michelle Fraden or Freighen.

2:00:25

Uh interestingly, she was at Sequoia.

2:00:27

She was on the FTX deal and then went over to OpenAI right before the Sam Alman firing.

2:00:32

And so it was just like chaos for a while, but it seems like she's settled in over at OpenAI and is doing well, but getting some push back from the Google folks.

2:00:40

Uh she says, "We hope you're doing well.

2:00:43

We wanted to follow up on our discussions over the past few weeks regarding using Google search API to help power chat GPT's search GPT prototype and search functionality in chat GPT.

2:00:51

As mentioned, we are currently testing a prototype search GPT and would and to learn more about how users and publishers use AI chat experiences coupled with real-time information and search-like features.

2:01:01

On July 25th, the day of our public announcement, we reached out to Google about using its search API.

2:01:06

We believe having multiple partners in particular Google's API would would enable us to provide a better product to users. So not just destroy you.

2:01:14

Um we also understand Google publicly offers its custom search JSON API to developers looking to incorporate Google search results into their products providing additional precedent for us to work together.

2:01:27

So basically asking like hey you have this service that you offer to a lot of people like can you offer it to us?

2:01:32

And they're like absolutely not.

2:01:34

Hey, do you mind letting the fox into the hen house? Summarize this email.

2:01:40

Fox sending the email to a chicken in the hen house.

2:01:46

Hey, I'd love to come in. I'd love to come in. Check out the house. I thought it might ask.

2:01:50

It seems like a great house.

2:01:52

I think my I think the other foxes in the fox house would like it. They like it. Fox.

2:02:02

Yeah, Google's like the the chickens over in the hen house are like I think it'd be kind of complicated.

2:02:11

Google said we're going to decline this because we believe that you are a wolf in sheep's clothing. That's hilarious. Well, yeah.

2:02:20

Cheers to Sundar for not letting the fox in the hen house and he lives to fight another day. Built his own Gemini.

2:02:28

Built his own apps and we're excited to talk to some folks.

2:02:31

built different built different.

2:02:31

Uh we're we're excited to talk to some folks over at Google about uh you know the the the hen house protection strategy. Yeah.

2:02:39

And the fox defense defense system.

2:02:41

The fox defense network, the Fox defense engine.

2:02:45

Should you talk about Coinbase? Oh, this is a rough one.

2:02:50

Marl Arrington is very upset with the Coinbase news.

2:02:52

Uh he says, "I'm a longtime investor and champion of Coinbase.

2:02:55

Something that has to be said though, this hack, which includes home addresses and account balances, will lead to people dying.

2:02:59

It's it probably already has.

2:03:01

The human cost denominated in misery is much larger than the $400 million or so they think it will actually cost the company to reimburse people.

2:03:09

The consequences to companies who do not adequately protect their customer information should include without limitation prison time for executives.

2:03:15

Uh very disappointed in Coinbase right now using the cheapest option for customer service has its price and Coinbase's customers will bear that cost.

2:03:22

And so yeah, so for context, uh if anybody was living under a data center, yeah, uh Coinbase effectively was offshoring their CX function.

2:03:35

Certain uh CX reps had access to consumer data, including home addresses and account balances, which they then effectively sold or were bribed into passing to nefarious groups, who then were threatening Coinbase, saying they were going to leak it.

2:03:50

Coinbase came out and said, you know, kick rocks.

2:03:51

We're going to pay uh basically a bounty that leads to the arrest of these people.

2:03:56

The issue obviously with home addresses and account balances being online is that uh uh bad actors could just find a home address if somebody with a high Bitcoin balance go there and say with a you know the threat of physical violence and say send me your Bitcoin right now or even worse abducted.

2:04:19

Uh and so Michael is basically saying that uh there's going to be a you know real real consequences here.

2:04:26

Coinbase came out and said, uh, anybody that sort of falls to various social engineering hacks due to the, uh, you know, they're basically going to reimburse them.

2:04:37

Apparently, they expect to spend quite a lot of money, hundreds of millions of dollars to to reimburse those people.

2:04:42

Uh, but Michael is saying that the the true cost is much higher.

2:04:46

And apparently, there had been just like a bunch of other issues here.

2:04:47

I mean, um, who knows what's actually going on.

2:04:52

Uh I'm sure that uh um I do wonder like how much of a cost center is customer service for Coinbase?

2:04:58

Like it seems like it's a big company, very profitable, has done very well.

2:05:01

Like if they doubled the cost of like the the expense, like would it get better?

2:05:07

Is is customer service really like a like a something that where spending more money on it is a lever or is this something that's like a little bit more intractable?

2:05:17

The other interesting thing is like I do wonder like the internet is pretty locked down.

2:05:23

I feel like like it it is like yes there are addresses that can be sold in the dark web but like it's pretty hard to just put up something that is illegal on the internet like like all the major platforms take it down.

2:05:38

Google shadowbans it uh the links get banned.

2:05:42

uh the browsers can be like like it's not but it's fairly easy for this information to just be like a file. Yes. An Excel file. Yes.

2:05:49

But at the same time, but then criminals can transact and sell multiple times at the same time like let's say that I'm like a you know the type of criminal that does that does just show up to a house and put a gun in someone's face.

2:06:01

No, the real the immediate risk is social engineering.

2:06:04

It's somebody calling you from a number that's spoofing Coinbase and saying, "Hey, I noticed some activity.

2:06:07

I need you to verify this. give me this code.

2:06:12

So that's the immediate risk.

2:06:12

Second order effect is that a criminal could just show up at somebody's house and demand payment. Yeah. Yeah. Yeah.

2:06:17

So, so the type of criminal that just shows up at people's houses, like how are they actually going to get this information?

2:06:25

Because they're not just going to Google it and it's not just going to show up there.

2:06:27

They're going to have to find it somewhere on like through some sort of hacker network.

2:06:30

Are they paying for this?

2:06:32

Like it feels like this information can be somewhat controlled.

2:06:37

Like credit card numbers leak all the time and and like the the global cyber security industry is pretty good at making these not just proliferate to like script kitties and like random criminals.

2:06:50

Like yes, there are organizations that try and like programmatically like run up a ton of charges and just like spam the network, but in terms of like these one-on-one dangerous interactions, like I'm not so sure that that it's it's going to happen like in mass.

2:07:05

Obviously, it'd be terrible if it didn't, and it's a very big deal, and it should be taken very seriously, but um but but but I do I do wonder if there's a way to to kind of like like not allow this information to proliferate.

2:07:18

Uh ironically, like what if what if it goes on the blockchain and it's just available?

2:07:22

It's like it's a very naughty issue with like, you know, censorship and uh and things.

2:07:26

Uh but Bology had a good response here to Arrington.

2:07:30

I don't have it clipped, but we'll have to talk to him about uh like the future of KYC.

2:07:33

His argument was that uh like part of the problem is that Coinbase is required to keep all this information when when they they might not actually need to if the government didn't require it.

2:07:45

And so it's this double-edged sword of the more information that you keep on someone, the more information that can be leaked.

2:07:52

Anyway, uh we'll dive into that more next week.

2:07:54

Uh for now, we have Austin Allred uh coming in to the studio. How you doing, Austin? Good to hear. Hey, good. How are you guys doing? What's going on? How are you? Yeah. Yeah. Hanging out. Well, yeah.

2:08:06

I wanted to have you on the show for a few reasons.

2:08:09

Obviously, there's a lot of stuff happening with um AI coding agents.

2:08:14

What's what's this whole AI thing? Yeah.

2:08:16

What's this whole AI thing? Whole AI thing. Haven't heard about it.

2:08:18

Um yeah, I mean maybe start with like introduction on yourself and and and and what uh what you're up to now, what you've done in the past, and then we can kind of go into uh the structure of uh of AI in the workplace, how people are learning to use AI, and then kind of some of the disruption that's potentially happening with AI tools actually replacing jobs. Yeah, for sure.

2:08:43

Um, so, uh, a little bit of background, co-founder of Bloomtech, which used to be known as Lambda School before a trademark lawsuit.

2:08:49

Um, it's technically still the same company.

2:08:51

Um, but about 18 months ago, we started talking to people about this new fangled AI thing and, you know, if it had any impact within engineering teams and frankly, I was pretty skeptical in the early days.

2:09:03

Uh, but we sent out a research team.

2:09:05

They looked at what everybody was doing.

2:09:07

And at the time it's not quite that way anymore, but there were a handful of people who had, you know, sat in the basement playing with LLMs, seeing if they could make them write real code for a couple of years.

2:09:19

And when we saw that, we were like, "Oh, wow.

2:09:21

This is actually working.

2:09:23

It's not mainstream at all yet."

2:09:24

And at the time, even less so.

2:09:24

Um, and then working with a couple of those companies that we were, so we started training engineers to, you know, use AI really well.

2:09:33

Um, and it's one of those things where there are no best practices.

2:09:37

Everybody's learning more every day.

2:09:39

The game completely changes every day, let alone every week.

2:09:41

Um, and then one of the companies we were working with brought us in to do a, you know, fully intensive 100h hour a week, find all the people with 98th percentile IQ and above, fly them into Austin, train them over the course of 10 weeks to be expert at using AI, um, and kind of staying on that that cutting edge, riding that wave up.

2:10:03

Um, and then they hire them on the other side and pay us to do that.

2:10:07

Now we're expanding that to more companies.

2:10:09

But I basically spent the past several months holed up in a little office with a bunch of sweaty people trying to figure out how to max out AI and figure out how to stay on the the cutting edge of using AI to build software.

2:10:23

Is the shape of that more like PM who's now able to write code or instantiate their ideas in code or is it software engineer who understands some basic programming that can then write more functional code or just be more performant like which which angle is it both or is one is one narrative driving more of like AI adoption?

2:10:45

Yeah, I'd say um it's more the latter.

2:10:49

So the more experience you have as an engineer, the more it sucks to start using AI because you know you're going from here to here as far as codew writing ability goes.

2:11:01

So in the early days of Gauntlet, we force everybody, you know, and the more senior you are, the more you hate it.

2:11:07

We say you can't write code manually.

2:11:09

We put software on everybody's computer that, you know, watches them and AI has to do all of the code writing for you.

2:11:16

And everybody thinks that that's the dumbest thing they've ever heard.

2:11:19

And they fight against you for a week, week and a half.

2:11:23

And then, you know, the more senior you are, the longer it takes for you to actually get to parody.

2:11:27

Um, but it's you, you'll never go back once you get there.

2:11:33

Um, it will, it has fundamentally changed the software industry forever in ways that people don't fully appreciate yet.

2:11:43

entire companies are going companies are going to have to rethink the way they do everything.

2:11:46

Um, and we're only seeing the very very very early stages of that, right?

2:11:52

Do you guys have your own kind of messaging internally with the students around uh the framing of like vibe coding because it maybe is it's like this cool viral phrase people. Yeah.

2:12:04

But it's kind of different but it's it's it's also different if you're doing it in this hyperintentional way.

2:12:13

trying to effectively get to the same quality and consistency as regular software engineering but just doing it in a super AI native way.

2:12:21

Yeah, I feel like that's muddied the waters quite a bit.

2:12:25

I mean the when I say vibe coding, most people imagine somebody who doesn't understand software kind of blindly trying to oneshot applications.

2:12:33

Uh which it's incredible that that works at all.

2:12:38

Um but when we think of you know whether you call it vibe coding or some people call started calling it superb building internally I don't know if that ever um made it out but using AI to write performant enterprisegrade bulletproof software um there's not a great distinction for that out in the market but we do view it very very differently.

2:12:58

There's a time and a place for both.

2:12:58

you know, if you're writing your own personal software or little internal tools like yeah, just blindly telling an LLM to do something might work.

2:13:06

But most of what our engineers are doing would work in, you know, very enterprise, very legacy, very big, large scale applications, you know, security is taken care of, the the formatting's done correctly.

2:13:21

It it's it's scalable and secure and all the things that software needs to be.

2:13:25

What are you seeing uh the most success with in terms of tools?

2:13:32

There's so many different options.

2:13:32

Every week there's a new tool, right?

2:13:33

You know, maybe they were using Devon at some point, then they're using Claude and then they're using uh now Codeex.

2:13:44

What what what's your kind of how how are people prioritizing?

2:13:46

Are they using things like Replet?

2:13:48

You know, what does that actually what does the stack look like?

2:13:52

Yeah, it it's not an exaggeration to say that it changes every week in a very fundamental way.

2:13:57

Um, so you know, I remember when Claude 37 came out, before Claude 37 came out, pretty much everybody was using Claude 35 as their daily driver.

2:14:08

Um, and I feel like Claude 37 was the first kind of breaking point where that became no longer true anymore.

2:14:15

Um, so before Claude had basically a monopoly.

2:14:18

If you were to walk around the gauntlet offices, um, and ask everybody what they were using, 95% was Claude 3. 5.

2:14:23

Um, with 37, it worked better for some people's workflows and worse for other people's workflows.

2:14:31

Um, Grock is gaining ground.

2:14:36

Um, it doesn't, you know, most people don't have API access to it yet.

2:14:38

So that's a, uh, you know, you have to use repo prompt or something to get it into your IDE.

2:14:45

So that's a you know separate thing.

2:14:47

Um I would say on average if for people who are working on really big really enterprise applications cloud 35 still seems to be winning out even over cloud 37.

2:14:57

Um and Gemini I think kind of changed the game. Um when was that?

2:15:00

Was that two weeks ago or a year ago?

2:15:02

I kind of lose track at this point.

2:15:04

Um but it has a much broader context window.

2:15:08

So if you have a small code base, you can get most of the code base inside the context window as opposed to cloud 35.

2:15:14

Um claw 35 is has proven to be better for more surgical stuff.

2:15:20

Cloud 37 is like cloud 35, but it kind of runs away and tries to do a lot.

2:15:24

Um so some people like that, some people don't.

2:15:27

Um, every engineer ends up with their very ordinate, well orchestrated workflow at Gauntlet and they, you know, they do things their own way.

2:15:35

Whether it's, um, you know, hey, I build all the front-end stuff really, really well and make sure that's all working and then I try to build a back end behind that or I diagram everything and plan everything out with a model like 03 and then I'm creating a checklist and pulling pieces off the checklist.

2:15:51

Every everybody has slightly different workflows, but uh it's on Twitter.

2:15:58

What What are you seeing on the other side on the on the on the hiring side?

2:16:01

Are companies expressing preferences already to you to some degree around, hey, you know, this is the stack we're using internally.

2:16:09

We want people to kind of be ramped already on on this set of tools or is it more kind of, you know, bring your own uh toolkit? Yeah.

2:16:21

So there's there's uh there's spectrum of course.

2:16:23

Um on one side I would say the companies that I I call them the companies that don't really get it yet.

2:16:31

They're like okay give me somebody who has five years of experience with Typescript and they understand a little bit of AI and we'll drop them into our traditional engineering org.

2:16:41

Um the companies that are on the other side of that spectrum are more actually this what was this entire team can now be one person.

2:16:49

Um and what language we use matters a whole lot less than how good they are at utilizing AI because anything that you don't know.

2:16:59

AI is better at writing code academically than any engineer you're going to talk to.

2:17:04

Um knowing how to get it to write the right code at the right time in the right place is where all the magic happens.

2:17:10

So on the other end of that spectrum, there are companies who are I mean the average company that we bring into Gauntlet and we show them what students are doing.

2:17:18

They walk away saying, "Oh my gosh, I need to rethink our hiring.

2:17:22

I need to rethink our entire strategy.

2:17:24

Our roadmap needs to change."

2:17:27

Uh it is a fundamentally different world when you're utilizing AI in the correct way.

2:17:33

And and then I guess even further on the other end of the spectrum, there are a lot of companies who tried it once and wholesale rejected it and aren't going to really use it or if an engineer does it at all, it's on their own time and against the will of the manager.

2:17:47

Um there's more of that than people would anticipate if you're spending all day on X like I am.

2:17:53

Um, but but yeah, there's a pretty broad spectrum and I go back and forth with frankly how much time I'm going to spend trying to convince people to do it all the right way or to experiment with new ways of building companies versus just okay, you know what, I'm going to go create a holding company and spin up 10 companies myself and show everybody what can happen.

2:18:13

Um, I I go back and forth on that, but but yeah, it's it's wildly wildly different. Mhm.

2:18:20

What is the what what's your read on the on the hiring market right now?

2:18:26

Obviously, there was big headlines.

2:18:27

Maybe it was last last week, I believe, midweek.

2:18:30

Microsoft that again, our our read on that situation was Microsoft has always taken taken the approach of kind of cutting the bottom 3%.

2:18:37

There still was a bunch of stories that came out from people being like, I've been here however many years working 12 hours a day.

2:18:46

Florida says they're not going to hire anyone.

2:18:47

hire anyone. that kind of went back on that like yeah overall hiring market temperature what's it looking like um I mean I think so first of all I agree with you that when there's a 3% riff that's rarely actually a riff that's performance management being done in as

2:19:04

kind of way as possible so you can give people severance um which if you're a company like Microsoft with as much cash as they have you probably should do it that way um that said generally say generally speaking what we're seeing is a if you're really really good at being an AI engineer. I mean, we had there

2:19:20

I mean, we had there were times when we had multiple billionaires wandering around our office with printed out offer letters for people begging them to come work for them.

2:19:30

Um, that's what it looks like if you're really good at AI.

2:19:33

Most companies are not really understanding that.

2:19:35

So, that's not everybody. Um, it is interesting.

2:19:41

I I think it's fair to say the junior engineering market is completely decimated. It's pretty well gone.

2:19:48

Um, even I mean if you're at at Stanford, maybe you can get an okay internship somewhere, but it is it's Armageddon to be. What does that mean?

2:19:58

Does that mean that like junior engineers should be vibe coding companies and starting more apps and trying to just like build lifestyle businesses almost like what what is your advice for people entering the workforce in software engineering capacities?

2:20:13

Yeah, that's a really good question. It's it's a tough one.

2:20:15

Um because the the way that I describe it is someone who's not very good at AI and AI engineering can still get roughly the output of a you know starting junior engineer out of AI.

2:20:29

And that's what a lot of companies are doing.

2:20:32

Instead of hiring a junior engineer, I'll just get you know the output that I need.

2:20:35

There's still and there always is crazy demand for senior and you know high-end and everything else.

2:20:42

Um I think it'll take a while to figure out how that plays out because the flip side of that is you don't get super senior engineers if you don't have junior engineers.

2:20:51

Um my and then you talk to the universities and mo the average university that I talk to is closer to ex explicitly and permanently forbidding any type of AI usage in the classroom than they are to interweaving it or adopting it in any material way.

2:21:13

Um because it's cheating.

2:21:16

It's so much better at writing code than you are that the professor has no idea if you know how to write code because the AI will write the code for you. Send him latest post.

2:21:29

I was at the I was at the park a couple days ago and uh was, you know, at the swings, you know, pushing the little guy.

2:21:37

Uh and there was another dad there who was a teacher and he said that that a high school teacher.

2:21:43

He said they basically now weight homework at effectively zero.

2:21:46

Like it's like kind of check the box.

2:21:49

It it maybe contributes to like a few percentage points of your grade and then tests are now just 100% because it's the only environment in which you can actually prove.

2:21:56

Would you advise people to learn to code now or has the idea of learning to code changed in some fundamental way?

2:22:04

So I still think I feel like I'm almost on an island on this.

2:22:09

I still think you should learn to use AI and if you want to use AI in software environments, you should learn to code.

2:22:17

Now, to me, that looks very different than it did five years ago.

2:22:22

Um, and our, you know, our learn to code uh business, we we've basically shut down. It no longer operates.

2:22:29

Um, I'm not interested in teaching people to code in the traditional way.

2:22:34

Um, now the flip side of that is normally when you would train people to code, you would start kind of close to the metal and at the bottom of the stack.

2:22:44

And so you would start with, okay, let's write binary and then we'll figure out how to write some, you know, low-level Java.

2:22:50

And eventually, a year and a half later, you start actually building programs that can do things that are interesting to you.

2:22:56

The right way to learn to code is actually in my mind the inverse where you start by building applications and then you figure out what's broken and what's missing and what you don't understand.

2:23:08

And the thing that is different about that than you know just a creating a different curriculum is AI understands that AI understands um you know so we had software that would basically watch what you were doing on your computer and from that we could derive what CS

2:23:26

principles you understood what CS principles you didn't understand we could have AI generate a custom curriculum that you know meets those guidelines specifically and we can fill any gaps in a way that I would have killed to a decade ago training people how to code. So I think you can actually

2:23:42

So I think you can actually just start by building stuff and then figure out how it works down and you know start at the top of the stack and slowly work your way down.

2:23:49

Um, and to what extent you get down the stack depends on what you want to do, right?

2:23:55

Not the average engineer today isn't, you know, playing with a Linux kernel or writing C.

2:24:00

They're, you know, writing JavaScript and they're building applications.

2:24:03

Um, to be at that level, you can probably get there relatively quickly.

2:24:07

Um, but I think I'll credit Martin Casado uh from A16Z with this.

2:24:11

um who said basically for to be a good engineer you've always wanted to understand one layer of the stack beneath what you were actually working on and I think that remains true. Yeah.

2:24:26

Uh give me your view on the AI software engineer market.

2:24:29

OpenAI now has like three offerings between you know 03 will just write code randomly for you. Codeex is a new product. They now own Windourf.

2:24:41

Um there's top down enterprise companies like Cognition and Devon.

2:24:45

There's bottoms up enterprise like Windsurf and uh Cursor.

2:24:50

Um how does this all play out?

2:24:52

What do you look what are you seeing in the like new this entirely new market of AI tooling?

2:24:57

Yeah, my my first point here is um we're in an incredibly lucky position because as a as a consumer, you don't have to predict who will win.

2:25:08

you can just figure out what's gaining momentum and latch on to the best.

2:25:12

Um, and so, you know, I have a lot of empathy for the companies trying to build those products and spend a billion dollars building a model and if something is 1% better, I'm switching to it tomorrow.

2:25:24

Um, so that's a great place to be in as a consumer. How is that playing out?

2:25:28

I think everybody's trying to fight similar battles.

2:25:30

Um, Cursor and Windsurf I think are uniquely positioned because they're focused primarily on the UX of the experience as opposed to the data.

2:25:41

Um, it's kind of in AI company's DNA to think that every problem is a data problem and they're generally speaking right.

2:25:48

Um, but where Windsor are getting better and better is okay.

2:25:53

It's really cool that there's this model.

2:25:54

I'd say on average that is the problem that AI companies have.

2:25:58

people have no idea what they can do with the models or how they should be using them.

2:26:04

I think that gets simplified over time the same way every product becomes, you know, I shouldn't have to know whether I want to use 03 mini or 40.

2:26:10

No normal human's ever going to care about the difference between those.

2:26:15

Um, so I think they'll just get better and better. Um, who wins? I I don't have a clue.

2:26:22

Trillion dollar question. Yeah, it is.

2:26:24

If you'd probably be a VC if you're in that game, but thanks. so much for stopping by.

2:26:29

We'd love to have you back.

2:26:30

This is uh this is great. Yeah. Keep us keep us posted. We'll talk to you soon. Cheers, Austin. Bye. Have a good one.

2:26:36

Next up, we have Jeff Morris Jr.

2:26:38

Junior himself coming in from Chapter One Ventures.

2:26:42

He's got some big news today. We'll let him back. Are you ready?

2:26:46

I'm gonna press this button. Okay. Welcome, Jeff. Jeff, how you doing? Welcome to the studio. Welcome. Come on in, guys. There he is. Our technology brother. Welcome to the show.

2:26:58

a uh a life life goal achieved.

2:26:58

I feel like I'm like let's go. Let's go.

2:27:04

No, I've seen I've seen this progress and I'm just such a big fan of of you both and um I won't flatter you for 20 minutes, but but really you guys are are killing it. It's awesome.

2:27:12

Well, I remember I think we got breakfast burritos in Santa Monica like right around the time that we were starting the show and uh you probably thought I was a little bit crazy for going full-time on a podcast, but uh but you you never you you always had faith.

2:27:27

I think you saw the the potential from the beginning.

2:27:30

So, it's great great to have have you on and um maybe give a quick intro and then uh I'll let you talk about the um the the news today and then we'll get into a bunch of other stuff. Yeah.

2:27:41

So um I've been an investor I guess for the past five years full-time started a firm called Chapter 1.

2:27:45

Um, we call ourselves the product fund mainly because all of our backgrounds are in product engineering, design, data science, um, and kind of take a product centric view towards what we invest in.

2:27:59

But prior to that, grew up in the area.

2:28:02

Um, probably most famously as an operator, was the VP of product at Tinder for a bunch of years.

2:28:06

Um, ran revenue over there kind of during like the hyperrowth period of 2015 to 2019 when the company was growing.

2:28:14

Uh and then now um now run the veter firm full-time and and yeah, we've uh evolved quite a bit.

2:28:23

Today we're sharing some of that news. Amazing.

2:28:25

Uh what's what's the news?

2:28:28

And uh we're we're not quite set up to have multiple guests on, but we'll have to have uh your new partner, well, new GP.

2:28:35

So why don't why don't you break down the news and and kind of talk about the evolution of the firm?

2:28:38

Yeah, so the big news today is Jameson Sadel is being promoted to a general partner.

2:28:43

Uh we started the firm in 2019 when there we go.

2:28:48

Uh you know I think it it kind of marks a a moment in time for us as a venture fund but also maybe more broadly in venture.

2:28:57

We started chapter one as a solo GP firm when being a solo GP I think was pretty uh new but also I'd call it trendy in 2019 and um the it's been just really interesting to see the peer group go in different directions.

2:29:11

of people who were still the GPS at the time.

2:29:13

Some have continued on that path.

2:29:16

Other people have gone back to operating and then I think there's another cohort of people who want to build partnerships and firms.

2:29:22

And so I think you know there's all these right ways to do venture whether it's um building a firm equal partnerships so GPS and there's this constant debate but I think we've um really been clear on what we want to become and today building a partnership is is obviously um what what we're announcing today. That's amazing.

2:29:45

Uh had this been your plan from the beginning?

2:29:48

uh like did you realize quickly that that you wanted investing at the firm to be a team sport versus something that you know was was more of this again solo endeavor at least like it was in the beginning.

2:30:01

beginning. I think a lot of us in 2019 like really like we had a plan but it wasn't um concept of a plan what we become and so I think I I realized in like 20 20 probably a year and a half two years in that I didn't want to just do it on my own and I think there's a lot of people

2:30:20

who are who are really happy with that lifestyle and um frankly it's probably a better financial decision just to do uh raise funds do it on your own but but I think to be competitive and to build a great product for founders, you likely can do a better job of that if you bring on partners and build a firm. And so

2:30:38

And so that was the goal in 2021.

2:30:41

It took a while to figure out who that person would be and also the right way to do it.

2:30:48

So a lot of people will tell you to kind of like promote people from within build, you know, you kind of like people buy into the culture.

2:30:56

know what what what what what you're doing and and you you kind of um uh elevate internally.

2:31:02

And other people say go like go find a superstar externally.

2:31:06

Um which which at times we also thought about doing so is you know like do you want to go bring bring in like a fancy GP spinning out or or maybe a a former CEO type and um those things are just really hard to to get right.

2:31:21

And so I think we um were really lucky where Jameson was doing an outstanding job internally, bought into what we were doing.

2:31:29

And um it's kind of like a head coaching search, right?

2:31:32

coaching search, right? you guys uh uh kind of in the TVPN model like it's really I think alluring to go hire like the big fancy um head coaching name but often times like promoting the person from within um is the best move because

2:31:50

they know they know your players your players trust them um and they know the system and I think I maybe like the Boston Celtics did that really well right and so I think it's really um uh I think that's kind of the decision most most firms have to make as as they're growing their teams. She's focused on I mean in her bio she

2:32:07

She's focused on I mean in her bio she says AI MLN infra how you guys making money Stargate there's 500 billion floating around it's the angle we were talking about just go set up a Chipotle there become a become a don't just set set up like a the the basically recreate the 2007 Chipotle in Abalene in Abalene massive opportunity there's 350 57 workers there.

2:32:35

Uh but I mean seriously like like obviously there's this AI boom.

2:32:39

A lot of the a lot of the trains have left the station.

2:32:42

I don't imagine you're participating in the next open AI round.

2:32:45

Um but but but how do you make money in the in the current AI boom or or is it uh or are we under like the next next trend?

2:32:56

Yeah, I think there's a couple angles like one is um location.

2:32:57

So there's Jameson lives in London.

2:33:00

Jameson lives in London. I think this is a popular narrative right now that all the like all the value is happening value creation is happening in the Bay Area and so obviously um I grew up in the Bay Area and I live in Los Angeles and we've made a decision to to focus kind of on the other most important areas in the in in tech and so um her view and it's my view being spending more time in London is that there's

2:33:25

actually a lot of talent um in the in the universities so Cambridge um Oxford etc and also So we we had a conference in London maybe two months ago and we had the founder of Granola, we had you know Anthropic, we had um all a lot of people from Jeep Mine and you plan to help those founders like escape the backwater that is the UK and the Europe because I can't imagine that they're going to try and build businesses there, right? That's crazy. That's crazy. Insane.

2:33:56

So like, but it makes sense because you go there, you give them money, and you tell them, "Here, go to San Francisco because then you have an actual shot." Is that the plan?

2:34:03

I think I think it's a it's a bit um like there's the there's like the the Project Europe movement, which obviously like love Harry and think that's important, but there's another view is like meeting founders where they are today.

2:34:18

founders where they are today. And often that that is hey we want to raise a seed round and and go you know I mean that's the story of anthropic and and granola right like like international founders who eventually came to the US there's always this question I have about like capital flows right Sweden's pretty

2:34:35

goated pretty good if you Nordics the Nordics there's something special there's always the capital flow like there are a lot of great LPs there that want to invest in American companies a lot of founders that want to come invest in America obviously there are some great companies in Europe where joking around. Um, but but but getting

2:34:49

Um, but but but getting that capital flow right so you don't just expatriate US dollars to underperforming countries is probably a risk you want to avoid.

2:34:58

But at the same time, it's probably really underpriced assets that are going to go find niche markets uh all over the world. And you know, who knows?

2:35:06

Maybe the next power law company comes from some bizarre country like do you have strong beliefs around the intersection of AI and crypto?

2:35:16

There's been a lot of different attempts and people attacking it from different angles.

2:35:20

You obviously have backed a bunch of crypto companies historically.

2:35:23

Uh yeah, how do we make money off of stable coins?

2:35:26

Ben Thompson was writing about this today.

2:35:28

How stable coins are going to web and I want to get on the action.

2:35:34

I'm super long stable coins, but they haven't moved at all. My portfolio is flat.

2:35:38

No, it is it is a tough thing where everyone's just so bullish on stables and yet and yet it's kind of unclear.

2:35:45

kind of unclear. Like a meme coin about stable coin USD on pumpf fun just USD coins potential potentially just one trillion there might be something there should be a dollar eventually no but but more seriously zero zero fiat backing it's just a completely completely it's just complete meat yeah unstable coins I think people have tried I think people have tried everything's been tried but

2:36:08

but seriously like stable coins uh what is exciting or again like uh with the bridge acquisition is that the end of the story or the I think you're seeing a lot of um at the application layer payments use applications are using stable coins um to enable their payments uh but there's not there hasn't been a lot of value acrruel to like net new um circle competitors are trying to compete with a tether or USDC. Yeah. And so stable Yeah.

2:36:35

And so stable coins are like the easiest thing to talk about at a AGM or LP conference because it's like okay like um I can kind of see that use case.

2:36:47

And so I think there's there's a shift to to emphasizing stable coins. Yeah.

2:36:51

Yeah, today I think the combination of like stable coins and agents starts to get like it does make sense, but it's it's it's gets pretty pretty blurry in like the current use cases of of of agents and and what's possible today.

2:37:08

Um I don't for what it's worth we don't um I have a strong belief on the intersection between crypto and AI today being um there's there aren't a lot of the the primary use case we like is giving open source developers a way to actually make money like I think that's a pretty clear thing that's been missing

2:37:28

if you look at any open source ecos ecosystem so you look at um I don't know like now research or different um like Tensor is trying to do this a little bit too, but the you know I I I I think the the worlds are are very separate today if you go spend time in the Bay Area and actually talk to the best AI researchers. Like there's nobody cares

2:37:49

Like there's nobody cares about crypto in the Bay Area for the most part.

2:37:53

Um and that's totally okay.

2:37:55

Like I kind of like that crypto's um still very weird and unliked by by most people.

2:38:01

Let's get Linus Torvalds a mega yacht.

2:38:03

Let's get Guido Van Rossom a private jet. This guy created Linux. Leguido created Python. Open source software. He didn't get comped.

2:38:12

Let's get him some stables. Immense immense value.

2:38:15

Let's get some greenbacks on chain. Exactly.

2:38:17

Uh I I want to talk about the uh the the the venture dynamics for uh these mega platform funds and smaller funds.

2:38:24

There was this uh there was this trend during the ZERP era where it seemed like the mega funds were just preempting everything.

2:38:34

You had the crossover investors and that made it pretty hard for early stage investors to kind of make the decision on whether or not to write proata checks into the into their earlier portfolio because they just wrote the seed check.

2:38:49

Maybe they did find the great company early, but then there was this pressure to say, "Hey, these companies are graduating to series A immediately."

2:38:56

You want to defend your ownership. Exactly. Exactly.

2:39:00

And so uh has has that dynamic changed or has the early stage market adjusted to spinning up growth vehicles so that they can uh ride that wave as the overall market gets hotter.

2:39:12

I mean, some of these some of these rounds for these AI companies, they get up uh series A and 100 million like happens pretty frequently now.

2:39:19

And how do you set yourself up for that as a manager?

2:39:24

I think that's probably the the hardest part of the 2019 or 2021 vintage, right? Yeah.

2:39:29

Um and you look back at at your follow on decisions and um at least speaking for ourselves, I think that was probably the the part of our investment we wish we could get back the most. Yeah.

2:39:40

Yeah. Uh, and so like today the markups are are just as crazy except there's there's obviously more revenue that you can you can kind of lean on to to underwrite the companies and um so it's actually it might be harder today because you have like before it's like

2:39:58

okay like Excel or index is marking up the company we did the seed you know we should maybe do the series A um out of principle now it's now it's like okay one of those firms is doing it plus they of, you know, 10 or $20 million in revenue. And so it's it's equally hard.

2:40:12

And so it's it's equally hard.

2:40:15

And I think for seed managers, um the growth fund thing for seed managers is pretty much done from what I've seen.

2:40:22

Like people aren't going and raising growth funds.

2:40:25

And if they do, um they have to have a very special reason why they're the seed fund who can do both.

2:40:31

So I think there's a lot more like we've seen SPD volume pick up again.

2:40:33

That's become a bigger part of the ecosystem.

2:40:38

And then I think a lot of a lot of LPs are looking for co-invest like that's always been the case but especially so today.

2:40:43

I think you see a lot of LPS coming to funds so they can co-invest and that's um I think a bigger trend today than when I started in 2019.

2:40:53

What's the sophistication level when they're looking at co-invest opportunities?

2:40:56

Are they saying, "Okay, I'm going to back an emerging manager, let's say a $30 million fund, and then I'm planning to do one or two co, you know, take one or two co-invest opportunities per fund and really just try to get into those winners, or is there more of a, you know, we just want broad direct exposure?"

2:41:15

And, you know, I'm curious what what you've seen or or what you're seeing broadly. Yeah, I think it's both.

2:41:22

Our preference is that you do broad exposure mainly because the odds of of doing one co-invest and having that be yeah the right company are just extremely low.

2:41:33

And so, um, we, you know, we don't, we don't do a ton of SPVS, but the messaging is always like, please don't overshoot on any single deal because, um, if you're doing a series B SPV, we all know the amount of risk event, the series B that and especially today, like that's even more pronounced.

2:41:58

And so, um, yeah, I think series B risk is more pronounced now.

2:42:04

I think if you look at the the amount of competition, the amount of change happening on a daily basis in in just technology advancements and then the pricing and size of these rounds, I would think if we look back two or three years from now, we'll we'll say the series B risk in 2025 was probably on par with the 2021 vintage would be my would be my guess. call it a bubble. It's crazy.

2:42:35

I mean, I I just see this having probably 60 plus different angel investments.

2:42:40

Now, I'll I'll get an update and see that a company's getting a markup and I'm like, I don't really think like they're there yet. They've made progress. Yeah.

2:42:50

Are they any closer to being a a business that truly has the used to be like preipo?

2:42:56

It used to be like this is an extremely solid business like within you know within sightelines of like true profitability and like I think we need I think we need a new stock exchange probably in Texas just called dogs which is like venture back ventureback companies that like have a little bit of revenue still got a lot of work to do.

2:43:17

Let's just get them out in the public market.

2:43:19

Let's let them trade at like two three million bucks.

2:43:20

Clear the prep stack and then maybe they 100x, right?

2:43:24

Maybe they figure it out, right?

2:43:26

Uh so anyways, the well what was that what was that spa uh that that holds there's that one there's that one destiny right now I don't know I don't know if you've seen this yeah the inverse destiny of the dogs the companies that are underperforming private markets you wrap all those into

2:43:41

a spack yeah investors can kind of sell to dogs take a little bit of a loss and you're like maybe there's one banger in here it's actually there's something there there's already businesses that do that that buy underperforming But yeah, this this this uh index Destiny, I don't know if you've seen it. They hold a

2:43:58

They hold a bunch of basically SP like they have like some exposure to SpaceX, but they traded like 10 times the underlying asset value.

2:44:05

Yeah, it's like I mean people want access to this to this stock, so they're going for it. Destiny Tech 100.

2:44:12

Anyway, um last question I have.

2:44:16

What's your updated thinking on incubations?

2:44:18

I know uh there's some stuff in the works that you probably can't announce, but um is that uh is part of bringing on a GP to to free you up to to kind of be able to spend a bit more time on on internal stuff?

2:44:34

That's been a a part of our stretch we want to expand on.

2:44:36

I think we're doing one incubation which I'd love to come on the show in like three months to announce and it's um partnering with a really big piece of IP like a a globally known piece of IP to build an application for that property.

2:44:51

And so I'm I'm I'm really excited about this because obviously building payroll software.

2:44:58

It's Mickey Mouse payroll software for sure.

2:44:59

It's honestly good execution and payroll you're at least a hundred million dollar company.

2:45:06

And so you slap some Mickey Mouse IP on that.

2:45:10

That's actually kind of genius. It's Mickey Mouse.

2:45:11

Just hyper commod like you go after these markets that are just hyper commoditized and just say like, "Yeah, we're partnering with with Disney.

2:45:19

It's just going to be Disney themed."

2:45:20

Oh, you don't want the BatmanVPN?

2:45:23

Why don't you want the Batman VPN?

2:45:25

You don't want the Dark Knight protecting you while you're browsing online.

2:45:29

Uh, you're going to go with what? Nord. What's Nord? Honestly, why not Jeff?

2:45:34

This is your new This is your new playbook. You live in LA.

2:45:35

Let's let's go on a road show.

2:45:38

Let's build some enterprise SAS meets legacy IP.

2:45:44

Buying it for pennies on the dollar.

2:45:44

You give him like 5% of the company. Yeah. Good to go. I do know.

2:45:49

I do know what Jeff What if Tony Stark built your ERP? Iron Man ERP. Stark ERP. Stark Enterprise ERP. This is the future.

2:45:59

Jeff could actually run this playbook. Uh, right.

2:46:03

Anyways, yeah, definitely definitely come back on.

2:46:06

Uh I I I think it's gonna really break the internet when you launch it.

2:46:10

So, I'm excited to see it. All right. Thanks, guys.

2:46:14

Congrats again to the whole team. Cheers.

2:46:16

Next up, we got Cliff Whitesman coming in from Speechify.

2:46:18

Uh great entrepreneur, great founder, good friend of mine.

2:46:22

Are we talking about business or lifting?

2:46:24

We're talking about lifting mostly.

2:46:25

So, he he's been on an absolute tear in the gym. He's super jacked. Uh no, it's not a joke. It's not a joke. Hit the Ashton Hall. Hit the Ashton Hall. Bring Br in.

2:46:35

Welcome to the stream, Cliff. Welcome to the stream.

2:46:40

Oh, we were expecting shirtless.

2:46:40

John said you were going to be shirtless. Just any TV.

2:46:43

You can do shirtless as well. Yeah.

2:46:45

No, no, we're making we're making TVPN stringers.

2:46:47

Uh yeah, I I I I worked out with him in in this building where we record this.

2:46:52

Uh he put up some fantastic numbers on the bench press, showed me a spreadsheet where he's uh what's it called? Thor 2. 0. That's right.

2:47:01

He's trying to become Thor 2. 0. now.

2:47:03

And so I've been telling him I'm trying to do Thor for 3.

2:47:04

0 and it's been getting under your skin.

2:47:07

But uh has a list in his gym. How's it been?

2:47:12

All the PRs of anyone who's ever worked out in that gym. Yeah. Yeah. Yeah.

2:47:16

We're we're getting there. Uh are you in LA now?

2:47:18

Well, where you at these days? I'm in LA. Nice.

2:47:23

I'm in LA in Studio City.

2:47:26

Uh but I'm about to we we move every four months or so.

2:47:29

So, I'm about to do SF, New York, Prague, back to New York, London, Rome, back to London, back to New York.

2:47:37

Those are all four month Those are all four months.

2:47:42

So, you're planning like years.

2:47:42

That's next That's the next 40 days. Okay.

2:47:45

So, bouncing around back in Florida for for a set period of time. Okay. Yeah.

2:47:50

So, so I mean, take us through the structure of the business.

2:47:52

Obviously, you can travel a lot, but uh I it's a fantastic story, and I just want to hear uh what you built, how you wound up there, and kind of the the the state of the union with Speechify. Yeah.

2:48:02

Give I I guess to give people a sense for the scale, which I think is important because typically people just think companies are big if enough VCs post about it, but you haven't raised very much money.

2:48:14

Speechify has over 500,000 five-star reviews, over 50 million users, Chrome extension of the year, app of the day.

2:48:22

Uh I'm just going to keep I'm going to hit this a couple times.

2:48:24

Um but uh but anyways, uh continue. Yeah. Yeah. Take us through it. Yeah. Happy to share more.

2:48:31

So I'm super dyslexic and I have ADHD.

2:48:33

So first, second, third, fourth grade, I had a really tough time learning how to read.

2:48:38

And when I was in right about to start college, I built this text to speech tool that would read out everything to me.

2:48:44

And um in high school, my mom used to read my summer reading books to me.

2:48:49

And just we didn't have time to finish the summer reading book for college.

2:48:51

So I cobbled together this thing that would read the stuff into my iPhone and then I'd listen to that on the planet. It worked.

2:48:58

And when I was in school, I studied renewable energy engineering at Brown.

2:49:01

Um I ended up building about 36 different products.

2:49:03

Everything from 3D printed skateboard brakes to iPhone apps and websites and payment systems. It's skateboard brakes. Skateboard brakes. I'll show you.

2:49:10

Tell Yeah, tell tell me tell me about that. You want to go slower.

2:49:15

So, um, if you ride a longboard or a skateboard and you go concrete hill, you don't want to stop by putting your face against the pavement.

2:49:23

So, what I did 3D printed the series of brakes that would attach to the back axle. Oh, wow.

2:49:29

That actually works really well. Ah, that's great. That's a good product. That That's perfect. Be huge on Shark Tank.

2:49:33

That's like the perfect product for that.

2:49:35

Could have been on the tank. Oh, you did. No way. That's the best.

2:49:37

I was considering going on a show, but it's like you end up sitting for like 48 hours in a trailer waiting to get called and another big opportunity that I had at the same time.

2:49:46

I was like, Shark Tank isn't worth it enough. Yeah.

2:49:48

So, at a certain point, I realized that software is much better.

2:49:52

So, I had built this kind of meme maker um on a flight from Logan Airport to SFO, published it, checked in on it 30, 40 days later, and it had 90,000 users.

2:50:02

And I was like, "Oh, software is way easier than injecting molding stuff and way easier than like boron doping silicon wafers."

2:50:10

Um, my thesis was on a more efficient solar cell that I was building. I was like, "All right."

2:50:15

And around this time, I read this paper about narrow applications of deep learning.

2:50:18

Uh, a bunch of academic papers.

2:50:20

One of them was called WaveNet about autogressive speech.

2:50:23

It came out of Deep Mind.

2:50:23

And I was like, "I can make a 100x better texttospech experience and a 100x better audiobooks experience."

2:50:32

and I wasn't sure what to work on.

2:50:32

I knew I didn't want to go get a job at Google or Palanteer or Meta.

2:50:36

Um, and I was like, "All right, well, I don't know what I want to do, so let me just write."

2:50:41

So, I wrote a 30-page paper about my worldviews.

2:50:43

And the conclusion was I wanted to be the person that I needed the most when I was a kid.

2:50:46

And the thing I really needed was someone to do my readings for me. Mhm.

2:50:48

Like, okay, I'm going to fully send it on this.

2:50:50

So, I convinced two of my professors to sponsor me to stay in school as a visiting scholar.

2:50:54

Basically, I guest taught classes.

2:50:56

I got to be a meal plan, live on campus, use the gym, but not pay tuition or do homework. Gym's important. Amazing.

2:51:00

I was going to do that uh work teaching computer science over the summer unlimited indefinitely until something took off.

2:51:08

And so 6 months in it took off.

2:51:10

Um and so now there's like 50 million people who use it.

2:51:13

Um but the goal is to make sure that reading is never a barrier for learning for anyone no matter what your background is.

2:51:18

So if you download the speechify iPhone app or Chrome extension or Mac app or Android app, it lets you take a picture of a physical book.

2:51:25

It reads um it gives you play buttons throughout the internet.

2:51:28

It's like the voice of the internet.

2:51:28

You click play, it reads.

2:51:29

Uh, and it coaches you to listen fast.

2:51:32

So, I listen to two audio books a week.

2:51:33

I've done that since I was 14.

2:51:34

Um, so I've read more than 1,800 books by listen.

2:51:36

You listen on like 3x or something? Yeah, I listen on 3x. So intense.

2:51:40

And so you can listen to the whole show in just one hour because it's a three hour. Yeah, it's one hour.

2:51:47

This is why streaming doesn't work for me.

2:51:49

It's got to be something that I can react. We're in RSS.

2:51:50

We're in We're in We're on YouTube.

2:51:53

So you can pull it in there. Listen.

2:51:54

That's exactly what I do.

2:51:56

I I show off the live streams, you know, 80%. Yeah. Yeah. Yeah. Catch up at the end.

2:52:01

Um, but look, I I learned English when I was 13.

2:52:03

So, in the beginning, I would listen at 75x speed, build up to 1. 5, 2x, 2. 5, 3x.

2:52:07

Um, and I'm obsessed with biographies, theology, philosophy, fantasy.

2:52:13

Um, and so in the beginning, nobody wanted to back a dyslexia education startup. AI was not yet hot.

2:52:21

Um the toughest point was um around 2017 we didn't have the level of engagement and retention that I wanted and there were a bunch of other sexy projects I could have done but from first principles my conclusion was I think that the trend that I most back is

2:52:37

narrow applications of deep learning to we call it generative AI uh from a supply side and from a demand side it's audio as a user interface and I was like the intersection of these two if I wanted to build something in that intersection it's speechify so I might as well stick with it. Um, I ended up

2:52:48

Um, I ended up putting a gigantic help button on the app.

2:52:52

It made like 20% of the screen real estate.

2:52:54

Bright red help last message us.

2:52:56

And if you clicked it, it just put you into an iMessage conversation with me.

2:52:59

So like I didn't need you to enable notifications. There was no intercom. There was no uh email.

2:53:02

And if I messaged you back and you didn't respond, I would FaceTime audio call you and a spreadsheet of all of our users.

2:53:08

And if for whatever reason you didn't use the product, I would call you every day until you used it to didn't use it.

2:53:13

And then the product became really good. That's amazing.

2:53:18

Um, we when we were 25, 18 of the folks who worked at Speechifi were previously either CEO CTO of VP of Engineering at their last company.

2:53:25

A lot of them were XYC founders or folks who exited their company after their series A. Nice tidbit on fitness.

2:53:30

12 of the original teammates got six packs within 10 weeks of joining.

2:53:34

Uh, and we had and we have had at least 17 people who gained 10 pounds of muscle um in the first 10 weeks. It's amazing.

2:53:45

Everyone back when we were here.

2:53:45

Um, and we got a big Airbnb in a different city every four months or so.

2:53:50

And so at the time we were in LA and I convinced one person to move from India, one person to move from Bulgaria, one person to move from Mexico, one person to move to San Francisco.

2:53:57

And I went and bought a bench press uh and I rented a truck, installed it in the apartment in my room.

2:54:02

Uh, and so we had me and Valentine were sleeping in and that's like the only kind of weights you need, right?

2:54:09

It's just a bench everything.

2:54:11

Um, and actually it gets even more deep because I went up to Marin to make sure that I was buying food for my parents because I didn't want them to go out during COVID.

2:54:19

And so two of the guys were in the apartment working out just body weight and they did a DEXA scan before and after and they found that they had like not gained any muscle even though they were working out like animals.

2:54:30

And I was like, "No, you guys need more weight."

2:54:32

And so I brought the bench press um and we had two guys sleeping in a bedroom in the living room that we partitioned with a window.

2:54:38

Two guys sleeping in the bed in the uh in the room and then one guy on an air mattress.

2:54:43

And that was like the best time ever. Like incredible time.

2:54:45

Um and then we found just very strong product market fit.

2:54:49

Ended up scaling really quickly.

2:54:50

We closed partnerships with all the top publishers to resell their audiobooks and ebooks.

2:54:52

My little brother Tyler started coding when he was seven building Dragon Ball Z websites.

2:54:59

When he was eight, he taught himself assembly to hack video games.

2:55:00

He went to exit for high school, skipped four and a half years of math, skipped five years of computer science, did Stanford as an undergrad, dropped out to run a cyber security company, went back to Stanford to do his masters in AI.

2:55:09

And I was like, hey, if you can help me build a model that uh will compile in 3x real time, has better quality than any API that's currently out there um and meets the following requirements, I will like do anything.

2:55:22

You know, I gave him a very nice compensation offer and he did it.

2:55:25

It took him about 10 months.

2:55:25

He joined Speechify 5 years in as head of AI.

2:55:26

So now we have a 40 person AI engineering team and we make the highest quality digital voices in the world.

2:55:31

We are the largest supplier of speech AI in the world for consumers.

2:55:34

Um and then so that's like level one, level two, level three.

2:55:38

Um the vibe now is we make everything multimodal.

2:55:40

So if you imagine that 5% of the population would read books for fun on their own.

2:55:44

Um and let's imagine 15% of the population would listen if you provided it as audio.

2:55:48

Now we're using text to video models to turn everything into like a full um audio and visual experience.

2:55:54

That's like super high level for speechify but happy to riff on anything.

2:55:57

How big is the company now? We're 176 people now. Wow, that's huge. 32 countries.

2:56:03

So five software engineers, 40 people on the engineering team and the rest is growth. Yeah.

2:56:09

Talk about talk about the things that venture has tried to make you that the venture industrial complex has maybe tried to make you do that you've rejected because in in a in an alternate reality you would have raised $500 million by now and you know and and you know decided not to do that and that's influenced a lot of decisions I'm sure.

2:56:30

So part of it is the fact that between the ages of 14 and 18 I read 461 books.

2:56:35

Um, a lot of them were about economics, philosophy, biographies.

2:56:42

Um, and I also listen to a lot of fantasy.

2:56:44

And when you listen to fantasy, you think, would I make the decision this character would have made in this point in time?

2:56:50

So, my favorite book is The Way of Kings by Brandon Sanderson.

2:56:53

I'm obsessed with this character named Kaledan.

2:56:54

I base my entire leadership style after Kaledan.

2:56:57

Um, and so when I was writing that 30-page paper to figure out what I wanted to do in my life, um, I became one of those lucky kids who when they were 21 figured out what they were put on earth to do.

2:57:09

And for me it was to solve dyslexia.

2:57:11

Um, and so I love what I do.

2:57:14

Like I wake up every morning early and I go to sleep very late.

2:57:17

I sleep very little because I just I can't get enough of working on this.

2:57:20

Um, and so I don't want to sell it.

2:57:23

I want to own as much of it as possible and I consider equity holy.

2:57:26

Um mind you I was a solo founder for five years working on speechify and again we had an amazing leadership team that uh blood sweat and tears we figured out how to hire.

2:57:35

Last year 50,000 people applied to work in the technical positions at speechify.

2:57:41

10,000 people took the asynchronous uh engineering challenges.

2:57:43

Um but we had like incredible offers for our series A from like literally the top firms. We said no. They doubled the offers.

2:57:51

We still said no because I didn't want to sell 20% of the company.

2:57:53

Um, and even with seed, we just picked folks who we thought were really good.

2:57:59

So, founders of Instagram, of Twitter, of Robin Hood, of you know, all these companies that I could learn from who did consumer subscription really well.

2:58:06

Um, and that was was key.

2:58:10

Um, so for example, Dylan Field from Figma taught me a ton very early on.

2:58:15

Um, and my nuance is if you take money from a fund, by and large, they have a fiduciary duty to their LPs.

2:58:23

If you take money from an individual, they have just a duty to themselves.

2:58:24

And so, you get a rolodex of someone who's really great, uh, but you're not beholden.

2:58:30

That being said, there are some businesses where it makes a lot of sense to raise money.

2:58:33

In our case, we have a very simple business and we understand the user extremely well and it's a very niche audience initially for people who have dyslexia, ADHD, low vision, autism, anxiety, concussion, second language learners, and then the entire productivity suite.

2:58:46

And at this point, we've worked on speech for eight and a half years.

2:58:49

So, it's really hard to catch up to the product innovations as well as the AI side.

2:58:52

Um, and so we are always talking to investors and don't get me wrong, we have raised money from, you know, funds as well.

2:58:59

Um, but we've just done it where we typically look for three things.

2:59:03

Number one, it's folks who either have an evergreen model, so they can ride with us because I'm going to be CEO of Speechify in 80 years.

2:59:10

Um, they have a history of philanthropy and education or an experience with dyslexia and their family.

2:59:17

And we think really highly of them.

2:59:19

They can see around corners that we can't.

2:59:20

Either they've done an amazing job taking multiple companies public um or they've backed founders in ways. So, you would go public.

2:59:26

We would go public but into the future not in the short term. Yeah.

2:59:31

What what are you thinking about competition from the hyperscalers, the big tech companies? Google IO is today.

2:59:39

Microsoft had their uh build keynote yesterday.

2:59:41

Uh it seems like an obvious place that uh competition could come from and yet uh on the product side a lot of the big tech companies just don't seem to be it iterating on the application layer side as fast as most people expected.

2:59:56

So what's that been like?

2:59:58

Has anything slowed you down over the past two years with competition from the big guys?

3:00:01

Oh, they've helped us a ton.

3:00:03

So I'll give you two huristics that are relevant to any founders then I'll give you the specifics on speechify.

3:00:06

The first one is in my personal opinion value occurs at the application layer more than it occurs at the model layer and models become commoditized.

3:00:14

Do you ever did you ever doubt yourself?

3:00:16

Did you ever think, oh, maybe value is going to acrue to the the model layer because there were there was like a there was like a 18month period where everyone was like models will do everything.

3:00:25

Well, I'm I'm giving you what my opinion is today and and the only reason I have a strong opinion is I've gone deeper into it than almost anyone has, right?

3:00:33

We bought millions of dollars of H100 GPUs.

3:00:38

We set them up in our own data center.

3:00:39

We have a 40 person team who does research and development on models.

3:00:41

So the second part I was going to say is but if you really want to be indestructible you got to own both the model and the data and the application layer.

3:00:49

Um and that's what made Chad GPT and OpenAI so successful is they had uh Brockman work on the userfacing side and they had Ilia work on the research side and if they just had the research they'd just be another lab.

3:01:00

It's application layer that made it incredible and uh so that's the first thing because you want to be able to control the quality and the cost as well as the user experience.

3:01:10

And then why did WhatsApp get acquired by Meta for $19 billion is because they had all the users.

3:01:14

And so you own the end relationship.

3:01:16

You want to have the phone number, the credit card, the email of that user, which we talk in a minute about kind of the changes that came with Apple.

3:01:22

Um, so that's the first thing.

3:01:25

The second thing is, and this is very unique, sorry, not unique, ubiquitous.

3:01:29

If you are Apple, Meta, Google, Netflix, there is no point in doing any project unless you think that project is going to make you one to hundred billion dollars per year in the long term. Mhm.

3:01:41

We had a point where we introduced translation into speechify. This is in 2020 2021.

3:01:44

And Mike Kger, who was the founder of Instagram, um CTO of Instagram, uh was like, "You should consider removing this feature."

3:01:50

And I'm like, "Are you kidding me?

3:01:51

Do you know how long I spent building this feature?"

3:01:54

But I have respect for Mike. So I took it out.

3:01:56

Install to trial increased, activation increased, users use the product more.

3:02:00

And I was like, I got to remove this. Why? Why?

3:02:01

Because it was distracting. Speech.

3:02:03

One thing to give you the best experience listening to something.

3:02:08

The second you add an extra button, users go down that path.

3:02:11

So, you know, from interesting has problem. But they had a problem.

3:02:13

You guys have such insane scale that you could immediately see something that again just removing some complexity.

3:02:22

Back back then we didn't have mad scale, but you could see it in 3 days of usage. like there's enough. No.

3:02:26

Um, so Clippy is horrible because Word is horrible because the design is terrible.

3:02:31

But you know who has amazing design? Google. Yeah.

3:02:35

One uh uh uh one text field.

3:02:35

Who else has amazing design? JPD.

3:02:38

One text field, one button.

3:02:40

And that's why they're successful.

3:02:42

If they had multiple buttons, they would be far less successful. Now, here's the key.

3:02:45

Text to speech is buried five levels deep in the menu.

3:02:49

When I was 19, I made a video on YouTube under an anonymous account because at the time I was embarrassed for people to learn I was dyslexic.

3:02:54

And it was called how to text to speech macree.

3:02:58

And if you search that phrase on Google even today, my video from when I was 19 is number one. Why?

3:03:02

Because it's so difficult to activate text to speech that you need 19-year-old Cliff to show you how to do it. And you can't pause. You can't play.

3:03:10

And by the way, if you want to listen like me, what I ended up doing in college is I would use the terminal to change the speed because the guey didn't go fast enough and it back down.

3:03:19

I had to restart my computer.

3:03:19

So I was restarting my MacBook multiple times a day when I was in college.

3:03:23

And then I was like, screw this.

3:03:24

I'm going to build a Mac app with keyboard shortcuts.

3:03:25

I'm going to build a thing that OCRs the screen.

3:03:28

And so my dream is for Google in mobile Chrome to add a gigantic play button that appears in every website that lets you listen to it. Why?

3:03:36

Because it will make everyone addicted to listening.

3:03:39

And then when they want to listen to a long PDF and they want to do it offline or they want to do it in my voice or Snoop Dogg voice or their own voice or they want to translate it to another language or they want to scan a physical document, Google doesn't cover that.

3:03:50

Apple doesn't cover that and people use multiple platforms.

3:03:51

So you need to use speechify and so speechify is the premium experience for text to speech.

3:03:57

What I need is for people to become educated about the idea that you can listen to stuff.

3:04:01

And so a lot of our work over the last eight years has been to educate the market.

3:04:05

And so there's really two goals.

3:04:06

Number one, build the most exceptional product like extreme.

3:04:11

We have four core principles.

3:04:11

Extreme product quality, leading with love, frugality, and speed.

3:04:13

So, we talk to users a lot.

3:04:15

We take really good care of each other.

3:04:16

We don't waste money on anything.

3:04:18

We build our own SAS tools and we move really fast.

3:04:20

We ship a lot to production.

3:04:22

Um, and the second goal is to educate people that, hey, you can practice being a fast listener.

3:04:27

And if you practice, anyone can do it.

3:04:28

My dad is 65 years old and he English is not his first language.

3:04:33

And the joke we have in my family is he wishes he had a mute button for me and I wish I had a speed up button for him. That's very sweet.

3:04:38

He now listens at two and a halfx speed to everything.

3:04:43

And so my life is better because my dad listens and talks faster now.

3:04:46

Uh but anyone can learn how to do this.

3:04:48

It's just a matter of practice.

3:04:50

When you come into first grade, no one expects you to be a good reader first, second, third, fourth, fifth, 12th grade.

3:04:55

We expect 12 years for you to become a good reader.

3:04:57

But you can become a good listener after listening to 15 audiobooks.

3:05:00

You're not going to be good on the first one.

3:05:02

You're not going to be good on the 10th one.

3:05:03

And being a good listener means three things.

3:05:04

Number one, you can do something else and listen at the same time.

3:05:07

Drive, cook, walk, whatever.

3:05:09

Number two, you can listen to more than 2x speed.

3:05:11

Number three, five weeks later, I ask you about the book, you have great retention of everything that you read in the book.

3:05:15

And it will not happen when you start. You have to practice.

3:05:18

But here's the thing, reading is a hack we invented, right?

3:05:20

24 characters on some dead trees.

3:05:24

Listening, we evolved to be good at listening, right?

3:05:25

Telling stories over the fire.

3:05:27

So if you were a bad listener, you were removed from the genetic pool.

3:05:30

If you're a bad reader, you were not removed from the genetic pool because otherwise I wouldn't be around.

3:05:33

Um, and so, uh, when you read, 30% of your brain is dedicated towards decoding. 70% is comprehending.

3:05:41

When you listen, like 3% is dedicated toward decoding, the rest is comprehending.

3:05:45

And so, you can understand a lot better.

3:05:47

And once you practice listening fast, you have that skill for life.

3:05:51

And especially if you have ADHD and you get distracted easily, the speed of listening matches the speed at which their mind is working.

3:05:57

And that helps people focus, especially in school. but also in work.

3:06:00

Um, and so that's the goal of the company is to just make that accessible to everybody.

3:06:04

And right now there's 450,000 audio books, but there's 100 million books.

3:06:06

We make all those accessible plus your emails, plus PDFs and everything else. It's amazing. Incredible. You are correct.

3:06:13

Uh, wish we had more time.

3:06:15

Yeah, we got to get a workout in soon. We will. Yeah.

3:06:17

Come by before you before you take off.

3:06:19

We got to make sure you break that PR. Yeah.

3:06:20

Yeah, I'm working on I I think I'm pretty close.

3:06:23

Uh, we'll talk to you soon, Cliff. Thanks. Thanks for coming on. You're the man. You got it. Bye. Bye.

3:06:29

Uh really bunch of VCs are going to reach out to him now.

3:06:34

He really has a fantastic business.

3:06:34

Uh let me tell you about Bezel. Go to getbzel. com.

3:06:38

Your bezel concierge is available now to source you any watch on the planet. Seriously, any watch. They got Rolexes.

3:06:46

They got the GMT Master 2, the Batman, the precursor to the Batman payroll suite or the Batman.

3:06:52

Uh you're really on to something there. I think it's good.

3:06:56

I think that's a feature. Disney payroll.

3:06:58

I mean, Logan Paul Logan Paul has a energy drink. Why not? Jake Paul has deodorant. Jake Paul has deodorant.

3:07:07

Why doesn't Batman have a ERP system? Why not? Or Shack Payroll. Shack payroll.

3:07:13

Shack payroll would be pretty good.

3:07:15

Um, before we bring in the Google folks, we should talk about Joe Weisenthal.

3:07:19

He says, "Your neighbor is online.

3:07:21

He thinks the bricks are creating a new currency that will soon replace the dollar.

3:07:24

He thinks Blackstone owns 60% of single family homes. He's long Cardano.

3:07:29

And this is a response to Buco Capital Bloke who says, "Your neighbor isn't online.

3:07:33

He doesn't think about tariffs.

3:07:35

His Honda CRV only has 80k miles on it. Good for another 100k."

3:07:40

He doesn't have an opinion on the rating of our sovereign debt.

3:07:41

Only the only flation he cares about is shrinkflation.

3:07:45

Yet his 401k goes up just the same.

3:07:45

I think the uh the Blackstone owning 60% of family homes is the funniest meme.

3:07:52

this whole idea of like, you know, oh, like Blackstone's the most powerful financial institution in the world just because they index everything.

3:07:58

Turning single family homes into pods. They are. Uh, all right.

3:08:03

Let's bring in you know what time it is. There they are. Huge day. How you doing? What's going on?

3:08:11

Congratulations on today. Uh, what's new?

3:08:14

Would you mind introducing yourselves and then just giving a quick breakdown of what's going on today over at Google? Yeah, I can go first. I'm Tulsi.

3:08:19

I lead the product team for Gemini. Cool.

3:08:21

And I'm work on some of the Gemini developer tools.

3:08:23

What are the top three announcements today?

3:08:26

What's the most exciting thing?

3:08:28

What is the the one takeaway for the audience?

3:08:34

Oo, I'll start with one takeaway.

3:08:34

I think the one takeaway is like Gemini, at least from my perspective, Gemini 2.

3:08:38

5 is just getting better.

3:08:40

Um, and it's not just getting better in terms of like the quality of the models themselves, but we're now bringing that capability everywhere.

3:08:49

So AI mode is coming to everyone which is kind of our advanced version of search.

3:08:53

Um Gemini is coming to glasses.

3:08:56

Um which is going to be amazing.

3:08:58

Uh Gemini is coming in even better ways to Gemini app.

3:09:02

Um with more weight limits and just more access and better features.

3:09:08

I was going to have to glasses.

3:09:10

Is this like a return of Google Glass? Is this a new Yeah.

3:09:12

So, so part of the issue for for us today is that we got to watch like 20 minutes of IO and then we started our own show.

3:09:19

So, we still have to catch up, but uh but yeah, and then and then it's good you mentioned rate limiting because uh I was worried I was going to have to spend, you know, this entire conversation moderating John asking for higher rate limits from you guys.

3:09:36

V2 need we'll make it happen.

3:09:38

Anyways, sorry Tulsi can continue. No, no.

3:09:41

I think it just it's it's like an exciting time for for us because I think the models continue to get better.

3:09:47

We released a new version of Gemini 2. 5 Flash today.

3:09:50

We introduced um like improved versions of 2.

3:09:53

5 Pro with the deep think kind of deeper reasoning version of the model.

3:09:58

And so I think the models are getting better, but it's also amazing to now see them actually really landing in Google products in in a very real way.

3:10:05

I don't know what what are your takeaways or top three are.

3:10:07

Yeah, I think my my TLDDR is Google's firing on all cylinders across every vertical, across every dimension.

3:10:14

And like IO is this like weird manifestation of this actually happening where you just like I'm like holy like even as someone who spends all day working with Tulsi, working with our teams and seeing this happen.

3:10:25

like IO is this crazy reminder of like just how much breath Google has in in the in the product areas and just how many people the model innovation touches and it's um yeah I think today's a celebration of all that hard work and and making all that happen. Yeah.

3:10:39

How do you think about the messaging from like the continuum of like consu Google's so big that like consumers will watch a developer focused keynote.

3:10:46

um some of the products that are launched are very consumerf facing but then there's obviously developers that want to build on top of these tools and then there's also stock analysts that are probably watching for how things are going on benchmarks or MAUs or any new data points and so um uh what are you focused on kind of messaging to the different audiences?

3:11:06

Is there a concept of speaking to multiple audiences or do you just try and laser focus it on on dev specifically?

3:11:11

No, I think we really do want to speak to multiple audiences.

3:11:14

I think you hit the nail on the head in like there are definitely experiences that we talk about here at IO that are very consumer focused.

3:11:20

We also know that developers are trying to build for consumers, right?

3:11:24

So when you showcase amazing consumer experiences, you are also able to speak to devs because you can speak to them about how you actually were able to build that experience and what they can do to build experiences like the ones that you're trying to build, right?

3:11:36

and those and so I think there's an art to try to figure out what are the right demos, what are the right stories that really can land cross audience and then we try to create sections that are maybe more targeted.

3:11:47

So for example, I spoke earlier today, Logan spoke in the developer keynote um really targeted at developers and what developers can do and then there are obviously sections around search or glasses or Gemini app that are much more targeted at consumers.

3:11:58

Walk straight off the dev demo stage right here. True.

3:12:03

Yeah, you were you said you could do two and you also said your keynote was starting at 1:30.

3:12:07

I was like, are you worried somebody might ask an an extra question?

3:12:10

Be in multiple places at once. It's a skill. This is amazing.

3:12:12

Um, uh, how important do you think those demos are?

3:12:17

Because, you know, we saw Chat GPT breakthrough with the, uh, the this the studio Giblly moment.

3:12:22

Everyone was memeing the Giblies.

3:12:23

memeing the Giblies. Uh I was recently using V2 to generate video of of cars with specific liveries and decals on them driving around and the result was fantastic and but it hasn't broken through in the same way even though I

3:12:38

feel like the model and just more importantly the accessibility because I was able to generate a video from from a Gemini chat box and I didn't have to go to a different app for that and I saw in the Apple App Store uh the the Gemini promoted post is basically like V2 is here. Um, and the output was so good

3:12:56

Um, and the output was so good that I showed it to somebody and they they said he was John had generated an image of a Lamborghini that had full TVPN branding on it and I showed it to somebody and they were like, "When did you do that?"

3:13:10

And I was like, and they were like, "That's right.

3:13:12

Was this how you made your promo video?

3:13:13

It was really just VO behind the scenes."

3:13:15

No, that's our that's our next promo video.

3:13:16

But actually, also on that note, you I think you probably haven't seen that part of the keynote yet, but we shipped VO3 today and it's awesome and it has audio. It's It's amazing. Oh, that's very cool.

3:13:28

Yeah, I was thinking about that because it was missing audio and I was like, this is the missing thing.

3:13:30

I need to go find another model and then figure out how to inference that or whatever.

3:13:33

Um, but yeah, I mean in terms of in terms of creating these like memeable moments or something that can go viral, is that something that's even thought of or is it more just like turn it loose and hopefully Well, yeah, I think the question is like, you know, you want to distribute the capability and the question is like is that Google's job or is it the developers building on your guys' platform to do that sort of last mile delivery to users? Yeah. Yeah.

3:13:56

mile delivery to users? Yeah. Yeah. I think the use case bit is is actually really important because I think despite like we're I feel like we're all sort of seeing everything that's happening and we're close to it, but like there's so many people who actually have no idea

3:14:10

and it's like every day I think we need like a tracker of like the number of new people entering the AI world every day, but like I imagine it's like some reasonable percentage of all people who are using these products and like you know most products are still a blank empty chat screen somewhere. So like

3:14:23

So like showing that use case of what's possible is super important.

3:14:27

And I feel like a lot of the demos from IO today were like pushing that frontier in a in a really meaningful way.

3:14:33

Um but I think also to answer the question of like whose responsibility is it?

3:14:38

Like I think Google we get to do both which is fun.

3:14:41

Like we get to go and build a lot of these use cases for developers and make it possible for them to, you know, bring that to their consumers, but then we also get to be like dog food and like really walk the walk and build those consumer experiences ourselves and find a way to make a a memeable moment. Yeah.

3:14:57

H how much of what is announced are more like showcases that could be built into other products long term?

3:15:03

Because I was looking at somebody was saying like Google Assistant has less fivestar ratings than Gemini, but you could imagine Gemini being like a demo for what's coming in Google Assistant, but it seems like Gemini's just gotten so big that it's like bigger than assistant now.

3:15:21

And so you have this weird thing where you're experimenting in this fun sandbox, but if you're really successful, then people just start using that and then you just move off of the other thing.

3:15:30

Is that is that like a deliberate part of the strategy or just like a natural outgrowth of like the modern AI development strategy that you put stuff out and if it goes really big you just let your winners ride.

3:15:40

I think it's a it's a good problem to have.

3:15:42

I think if we're in a world where the experiments we're trying or the new kind of surfaces we're building or the new types of use cases we're building actually really resonate with users and then we're in a place where we actually want to build those and scale those out.

3:15:56

I think what we're trying to do, so at the end of the day, what we want to build is products that people love and really want to use, right?

3:16:01

And I think there's three ways you can do that.

3:16:03

You can try to build new features into your existing products.

3:16:06

I think in a lot of cases that makes sense, right?

3:16:08

So, for example, like AI mode on search today. Yeah.

3:16:12

Um, but in some cases, if you do too much of that, you bloat the existing product.

3:16:16

it becomes harder to use and understand versus being able to create a new surface where you actually can really rethink the UX, really rethink how users are supposed to think about the product.

3:16:25

And then you can look at that and say, well, does this really make sense as a standalone or does it actually uh merge with other ways that users are using different products, right?

3:16:35

And actually fits into their existing workflows in some way.

3:16:37

And so I don't think we want to pigeon holes ourselves into just, hey, we need to consistently build on one product, right?

3:16:44

We want to give ourselves the flexibility to create ways for users to build new mental models and then figure out how those things come together.

3:16:50

And I think this search example is actually a great one because you if you saw we announced deep research um in the Gemini app came out I think all the way back in December and it was like pioneered the deep research category and Google launched that in in the Gemini app and people love it.

3:17:04

It's it's a crazy product experience.

3:17:05

Um and then like that's now a like I think the the you know the feedback was this is an incredible experience.

3:17:11

Not everyone in the world is a Gemini app customer.

3:17:13

Um there's a whole lot of people in search and like now I think it's called deep search is available as part of AI mode inside of just like basic Google search which is going to reach hopefully even a many many millions of more users than are using the couple billion couple billion. Yeah. Um just a just a B. Talk about MCP.

3:17:33

Was there was there uh like a sigh of relief that Google's leaning into MCP?

3:17:39

Um, was it ever a question or was it something that was like pretty easily adopted across Google?

3:17:44

I think it's been fascinating to see like how quickly the world has like jump I feel like with standards like people I feel like are often like very slow moving to be like do we adopt this?

3:17:55

There's like a whole you know many year battles with different standards competing and I feel like everyone was just like MCP came out everyone's like okay we'll use MCP. XQCD. It's that XQCD meme. There's 10 standards. We need another ST.

3:18:08

We need just need one standard.

3:18:09

Now there's 11 standards.

3:18:11

But in this case, we actually got the good ending.

3:18:12

It's important though like I think developers are building stuff.

3:18:16

I think in this moment where like you know the world is still trying to figure out how to build robust agentic systems like having people not spend all the time just rewriting the same framework 50 times over like is I think is like actually a net benefit for the world.

3:18:30

So, and I think the the team that built anthrop uh that built MCP at Enthropic did a really really good job and it was a smart move by them to like actually make it an open standard which um they could have easily not done and I think it's it's the right thing for the ecosystem. Yeah.

3:18:43

Puts developers first and I think the right way. Yeah.

3:18:45

Um what what is the overall messaging to developers?

3:18:50

It seemed like Satcha on stage at at Microsoft's Build was very much like we host every single model.

3:18:57

We'll we'll give you DeepSeek if you want. We're partnering with X.

3:18:58

Um Google in my mind has always just been like look, we invented the transformer.

3:19:03

We have the biggest like best models that we're we're domining the paro curve.

3:19:06

Uh are are you guys seen more as like a one-stop shop or do you think there'll be more uh model agnosticism in the future with regard to GCP?

3:19:18

I think GCP already has this.

3:19:18

So GC like we do like obviously you know Tulsi and the team do a great job with Gemini models.

3:19:25

We want the world to use Gemini, but GCP does actually have like I think it's over 175 different models that are available through the model garden.

3:19:30

You can get anthropics models.

3:19:32

You can get a lot of other models.

3:19:34

So I think the and like for customers on cloud like that makes perfect sense.

3:19:38

They need that like that's a requirement like the world doesn't want to you know be locked into a single model.

3:19:43

Um but I think like we get to see these like really really interesting experiences that can only be built on into Gemini because of the like stack of of Google's model innovation.

3:19:54

Um, so we're going to keep pushing on on Gemini models, too. That's the goal. Yeah.

3:19:58

And then on the uh uh developers that might be working with Google, I imagine that the conversation is a little more nuanced uh if you're actually scaling a growing app on on on Google infrastructure.

3:20:09

like I ran into a rate limit as a random user, but uh if you're if you're actually thinking about scaling a startup, uh unless you're, you know, thinking about building Stargate, uh you're probably able to scale pretty pretty efficiently.

3:20:25

But have you been seeing companies run into um any sort of like scaling or rate limits or challenges as they kind of grow on Google? Email me.

3:20:32

My email my email's online.

3:20:34

I'm on Twitter 14 hours a day.

3:20:37

So, whatever you need, email me.

3:20:39

We'll make sure it doesn't happen.

3:20:39

I think in general like this is like as far as like a class of problems that developers have in this moment.

3:20:44

I think that like very real like I I hear this and see people all the time.

3:20:47

The challenge is like compute is like a it's like literally a currency almost.

3:20:52

So like we have to make sure we get the right compute to the right people who are actually building stuff and like it is a it's not an easy process to do that.

3:21:00

But I think if people need stuff please reach out to the team actually reach out to Tulsi too.

3:21:04

Uh we'll get yeah spread the spread the love you know.

3:21:08

Don't just send your don't just don't just create CX tickets for Logan.

3:21:13

Uh maybe talk talk about a couple of the the other kind of product level announcements you guys had.

3:21:19

Jules uh an asynchronous coding agent.

3:21:22

You want to speak to that a little bit?

3:21:23

I'm curious to get your Yeah, we're really excited about Jules.

3:21:27