0:06
[Music] Heat. Heat. Heat. Heat. [Music] N.
Search this video
0:46
Heat [Music] [Applause] [Music] up [Music] here. [Music] Heat. Heat. [Music] Heat. [Music] Heat. Heat. Heat. [Music] Heat. Heat. [Music] N. Heat. Heat.
2:42
[Music] [Applause] [Music] We came to this [Music] world to reach the stars.
3:21
We came to this [Music] world to see a future.
3:30
[Music] We came to this [Music] world to reach the thoughts to shape our future.
3:47
We change to feel the new [Applause] [Music] [Music] [Applause] [Music] Double kill.
4:24
[Music] If you let [Music] me feel [Music] [Music] You're watching TV.
4:58
Today is Wednesday, May 21st, 2025.
5:01
We are live from the Temple of Technology, the fortress of finance, the capital of capital.
5:06
I still have to send the call-in post letting everyone know what we're talking about today.
5:12
And that's why I'm a little bit behind on You got a little bit of a lag, but if not, John, we can just all get a bit of a laugh out of you. No, no, no, no.
5:21
You're No one's laughing at me.
5:21
The studio audience is does not find this funny.
5:25
Um, there's nothing funny about what's going on in the news. We got Open AI.
5:30
They bought Johnny IV's secret AI company, IO, for $6. 5 billion.
5:33
We got reactions to Google IO.
5:36
I have been calling everyone I know who works at Google to try and get me more V3 credits because I maxed it out immediately. Uh it's fantastic.
5:47
The videos are extremely impressive.
5:49
Uh there's still crazy hallucinations.
5:52
We'll dig into it, but it's a lot of fun.
5:53
Highly recommend you go and fork over the $500 a month. I think it is only 50.
5:58
You can get it for 50% off right now. I'm now Yeah, 50% off.
6:03
So I think they hit me with three 250 but I get 2. 5 flash.
6:05
I'm on Gemini Ultra, but I can't do video because I maxed it out.
6:12
Um, but the results were fantastic.
6:14
I posted some earlier this morning.
6:16
Uh, Jordy posted one of him fighting uh 100 gorillas and winning possible with artificial intelligence. Yeah.
6:23
You simply wouldn't be able to find out the answer to that. Exactly.
6:24
Of you know, a 100 gorillas versus one tech podcaster.
6:29
You wouldn't you would need AI to really simulate that kind of conflict.
6:34
Then we're talking about uh Europe trying to get into the technology game.
6:39
I I think of technology as I think of technology as as like a uniquely American thing, but apparently they're trying to do it over there.
6:49
I tend to think of invention as an American invention.
6:52
And so I've never heard of any technology coming out of Europe, but I guess people are trying to make it happen.
6:58
So, you know, anything's possible in the in the future of AI abundance. It's possible. We love our Nordic.
7:04
We love our Nordic entrepreneurs. They're goated. Yes.
7:05
And we love one Italian entrepreneur in particular, John Furantino.
7:10
Of course, of course, Volulta, of Fantino label, of Moon Pals, of Gravity Blanket.
7:15
You know, he's probably the best Italian entrepreneur in history.
7:20
Definitely in the conversation, right?
7:22
Definitely in the conversation, John.
7:23
Anyway, let's kick it off with uh let's also pull up the guest list, tell you who's calling in. We got Eric Prince.
7:28
We got David Senra from Founders Podcast.
7:29
We got Scott Woo from Cognition breaking down that question that I've been digging into.
7:35
How does the AI coding market uh landscape evolve now?
7:38
Uh we got top down enterprise approaches, bottom up enterprise approaches.
7:43
Windsurf's been acquired.
7:45
Cursors on a tear raising at 9 billion.
7:47
We got codeex and jewels from Google.
7:50
Now everyone has an AI uh companion, an AI coding bot, a companion for the software.
7:56
You're going to have you're going to have coding agents that code you your own AI companion. That's the future.
8:03
We're the real the real test for Jules is can Google's new AI uh AI coding agent build oneshot me Google reader beloved Google product sunset. I want one prompt. Build me Google.
8:17
I was I was blown away by by everything that Google announced yesterday and uh I was kind of laughing a little bit because one of Google's agencies was was interested in in partnering with us. Oh yeah.
8:27
And the onboarding just to get in the conversation with them has been so intense that I just haven't done it for for weeks.
8:39
And so it'd be cool if they build an an agent that could sort of uh build, you know, an onboarding flow that that didn't involve a bunch of Excel sheets. That'd be great.
8:47
That's a anyway longer conversation.
8:49
Uh let's go to the the the breaking news.
8:52
Open AAI is buying buying Johnny IV's secret AI company IO for 6.
8:54
5 billion uh billion dollars.
8:57
Uh although it is in stock and they will develop a slew of devices.
9:02
Uh Iive and his return to hardware and love form will oversee OpenAI design all the details including interviews with IV, Sam Alman and others from Mark German in Bloomberg.
9:11
I mean what he is assembling the Avengers all over again.
9:16
It's crazy like the the early OpenAI team was insane. Yeah.
9:19
And now he's putting together Fiji Simo, you got Kevin Wheel, you got Johnny IV now.
9:27
There's just more and more people coming. 9minute video.
9:28
The video basically does a lot is right.
9:31
I didn't get to lot of back and forth.
9:35
I summarized it with Chad GBT.
9:37
Basically said Johnny's amazing, Sam's amazing, and they're cooking.
9:41
That was like the three big takeaways from the video.
9:43
So, if you don't have time, if you have time to watch this three-hour podcast, but you don't have time to watch that 9minute video, uh, that's kind of the I really did like the cinematography for what is essentially a 9minute podcast clip.
9:55
Feels like you're a barista, but it's Yeah.
9:57
And that kind of makes you think maybe they're signaling that's the last job. Maybe, maybe, maybe.
10:01
Venture capital and being a barista.
10:04
It'll be very interesting to see what they actually launch.
10:07
Anyway, let's dig into the Mark German article.
10:08
Uh, Bill Gurley immediately within minutes of Mark German posing this cash or stock. What was the deal? I want to know. I want to know.
10:17
Uh, and Mark German says five billion in stock.
10:19
Open AAI bought an initial 23% in Q4 for about 6. 5 billion in total.
10:24
Uh, Patty's in there saying, I don't want an OpenAI phone.
10:27
I want an OpenAI brain chip.
10:29
I want the 40 hallucination straight to the brain. Love Patty. Hope he's doing well.
10:34
Um anyway, let's let's dig into this article and break it all down.
10:38
So from Mark German in Bloomberg, OpenAI to buy AI startup from Apple veteran Johnny IV in $6. 5 billion deal.
10:45
They will acquire the AI device startup co-founded by Apple Inc.
10:50
veteran Johnny IV in a nearly all stock deal for $6. 5 billion.
10:52
The purchase, which is the largest in OpenAI's history, will provide the company with a dedicated unit for Divi for developing AI powered devices.
11:01
And that makes a ton of sense given that they are subject to the platforms.
11:06
Google obviously is Android.
11:06
Tons and tons of AI products as we saw yesterday.
11:10
We're going to go into some of the Google IO reactions and what Ben Thompson's saying on strategy about uh about Google IO.
11:15
Um and then Apple Apple's obviously a little bit behind the ball with Apple intelligence, but they can charge 30% on everything.
11:21
They can, you know, gate things and they certainly can't let OpenAI have free reign of the OS as they'd certainly like to have.
11:30
And so uh not a re not an unreasonable decision although many companies have tried and failed to break into hardware.
11:36
Facebook famously with the uh with the Facebook phone which I believe was a project that was run by Chim Paul.
11:41
The interesting thing here is that so openai is not acquiring love from which is the con sort of consulting agency by Johnny. Yeah.
11:48
And but love will work continue to work on IO internally at OpenAI. Okay.
11:54
and uh not sorry not internally but IOIO is now part of OpenAI.
11:59
So the big question for me is $6 half billion is a big number. Yep.
12:04
They must have made some meaningful amount of progress otherwise they could just say to to love from can we just pay you you know $500 million a year to develop a device ourselves. Yeah.
12:14
And so in the video, the thing that was notable to me is that Sam has an early version of whatever their initial device is that he says he's blown away by and uh they've gone out and said they're making a family of products as well. Makes sense.
12:29
I mean, at the same time, if you flip it around, what does it take to get Johnny IV to join your startup?
12:34
Is it close to 2% equity? Probably.
12:37
Couple couple points at least. A couple points, right?
12:42
I mean, that's not unreasonable.
12:43
And so you run that math and all of a sudden it's like, well, we're taking we're paying for this mostly in stock. It's all stock deal. 6.
12:48
5 billion latest valuation is what? 350 300 330 330.
12:51
And so it's it's basically 2% of the company of OpenAI is going to Johnny and his team.
12:59
He's going to slice that up.
13:01
Probably owns a bunch of it.
13:01
And so what would it take to get someone to be to get Johnny IV, the legendary designer, to be their head of design.
13:07
You know, if you were any other company, you'd be like, yeah, one 2% probably, right? Yeah. senior executive. Yeah, exactly.
13:14
Like like the the numbers they're all they all have four extra zeros behind them, but but they but they all make sense when you when you rationalize it that way.
13:22
Anyway, um so there's going to be this dedicated unit for developing AI powered devices.
13:26
Acquiring the secret startup secretive startup named IO will also secure the device the services of IV and other former Apple designers who were behind iconic products such as the iPhone. Uh here's a quote.
13:35
I have a growing sense that everything I've learned over the last 30 years has led me to this place and to this moment. Very poetic.
13:43
Uh I've said in a joint interview with OpenAI Chief Executive Officer Sam Alman, "It's a relationship and a way of working together that I think is going to yield products and products and products."
13:53
For the Britishborn designer, the move marks a high-profile return to a consumer technology industry he helped pioneer.
14:00
working for years alongside Steve Jobs.
14:02
He crafted the look and feel of the modern smartphone in addition to the iPod, the iPad, the Apple Watch.
14:06
He left and Johnny specifically called out that these are decades old devices, right?
14:13
And so his point here is what what comes next, right?
14:16
What is the next platform? So he left it in 2019.
14:18
Obviously Apple was working on vision pro for more than five years, right?
14:22
So he must have seen that project at some point and then been like I got I got to leave.
14:27
I got here resignation letter. No. Uh I don't know.
14:33
I mean, obviously it has it has touches.
14:35
And the crazy thing, there's this picture of the two of them here, which uh Johnny unfortunately fails the green line test.
14:40
He did in this case did that for for us.
14:43
Tagged us in it actually. Yeah.
14:45
But but the crazy thing here would be if if Johnny is actually wearing the first device, right? We don't know.
14:53
We don't know the form factor, right?
14:53
It could be something like a pendant. It could be glasses. Maybe something else.
14:59
Uh I'm just speculating, but uh I wouldn't be surprised if there's Easter eggs, but I mean the green line test here makes sense.
15:06
It's like Sam's the CEO and Johnny works for the CEO now. So yeah.
15:09
So it's hard to say he failed maybe in represented the true relationship of the business partnership, right?
15:13
Always always fail the green line test for the boss man. Yeah. Yeah.
15:17
You don't want to you don't want to mo your boss in a photo. That's a disaster. Yeah.
15:20
IV was once described by Jobs as his spiritual partner and his new stint designing rival technology products could be seen as a bad omen for Apple, a company already struggling to compete in AI.
15:30
In the interview, Altman said Jobs would be damn proud of IV's latest move.
15:34
Open AI is going to create a product at the level of quality that has never happened before in consumer hardware.
15:40
Alman said AI is such a big leap forward in terms of what people can do that it needs a new kind of computing form factor to get the maximum potential out of it.
15:48
He said, "Not saying what form factor that is, but could it be a rectangle with a screen?"
15:54
The critique of the video, if it was a a beautiful conversation that shared almost nothing other than they clearly admire each other and they're excited about what they're building.
16:06
It's just two guys hanging out yapping basically. It's ridiculous.
16:09
Why would anyone do that?
16:11
Why would anyone publish multiple minutes of just two guys talking to each other? This is crazy.
16:16
I love I love at the beginning of the video, you know, they're clearly walking to meet at this coffee shop or bar, whatever they're meeting at.
16:22
And to me, the idea of the two of them just walking around San Francisco seems probably hard to go that far without, you know, I was thinking about that.
16:33
I was like, if I saw Johnny just cross street, he's jaywalking, right? Yeah.
16:39
Just you can just do things. Yep. You can just do things. Jaywalking. Yeah.
16:45
So, they just walk in and uh yeah, full camera crew. Very cinematic shot. Really, really cool. Beautiful.
16:50
I called it, by the way, Ben just broke the news to us in the chat that there was an earthquake in LA.
16:56
I was sitting here at the table a minute before a minute before a minute before we went live and I'm like, different.
17:01
I was like, there's an earthquake happening right now.
17:04
John's like, "No, I'm just banging the table. I'm scared.
17:06
Your water your water is shaking."
17:09
I was the the the This is show biz, John.
17:12
the show was going to go on.
17:14
We were not going to stop the show, but there I was a couple imagining things.
17:19
Um, anyways, the the other thing here, uh, OpenAI had acquired a 23% stake in IO last year. Yeah.
17:27
So, already pretty major shareholder and, um, separately, OpenAI startup fund also invested in Ives company the last at the LA at that time, billionaire philanthropist Lauren Pal. There have been rumors. Do you remember?
17:39
Do you remember like rumors?
17:39
I I feel like it was about a year ago that Sam and Johnny were like in the Middle East pitching this like device product.
17:45
So there had been a bunch of rumors um but never a ton of clarity.
17:50
I mean Sam just understands the full implication of the AI stack and is and is interested in in in having a play in all of them like the helon.
18:00
He's thinking about fusion and energy at the super super long tail.
18:03
He he was talking about chip fabrication yesterday.
18:07
He's obviously doing uh data center buildouts now as well as the product level and the foundation level.
18:15
It's like yeah if the oh where does the value acrew the model layer or the application layer?
18:19
It's like it doesn't matter if it acquire if it if it if the value acrews to the the foundation model layer, the application layer, the energy layer, the data center layer, the chip fabrication layer, the device layer, or the energy level like he he wins because he's got he's got full stack. He's verticalizing.
18:37
Yeah, he's verticalizing.
18:39
The deal is expected to complete to be completed this summer pending regulatory approvals.
18:43
The takeover of IO will And by the way, if this gets blocked, I'm taking to the streets.
18:48
I'm going to be I'm going to be out there.
18:51
Two guys can't even get a coffee and build together anymore. Yeah. Yeah. It's ridiculous.
18:55
I do wonder how Elon's going to respond to this.
18:57
He's probably going to be like, "This is ridiculous.
18:59
I don't like this at all." It's very annoying.
19:01
Yeah, it's very annoying for for Elon.
19:07
for because to be one of the most powerful wealthy people in the world and have it out for somebody and be so publicly aggressive towards them but not seem to be able to actually, you know, he's consistently trying to, you know, Sam's biking by, Elon's trying to throw like a broom through the through the wheel, you know. Yep.
19:30
And it's just the wheel's breaking the broom.
19:35
I feel like maybe Elon needs to pick up somebody, some some iconic designer.
19:38
Uh who uh is uh is Penn and Fina available or something?
19:49
What are they up to these days?
19:49
I mean, we've talked about like Mansori for Tesla, but maybe he needs to get someone to design some some X hardware.
19:55
to design some some X hardware. The funny thing is that like there have been these fake but I would be interested I would be interested to it's it's very possible Elon has a different thesis around hardware and AI right it might not his point of view who knows could be it's Neurolink is all that matters like you guys are focused
20:15
on the wrong form factor I'm sort of playing playing 10 20 years in the future you guys can you guys can make cute glasses I'm going to be in your brain yeah like that that's Elon's lawn game is He's like going to make the best and like Sam is gonna have to at some point be like, "Okay, yes, the chat app is available for Neuralink." And I got
20:34
And I got to get the chip.
20:36
And then Elon's like, "Gotcha. Gotcha.
20:37
You let the you let the fox into the hen house. Fox into the hen house.
20:42
You let the chip in your brain. Yep.
20:45
And it's just like imagine he's like, you know, Sam Sam has endured, you know, having Elon, you know, comment the like crying emoji, but imagine Elon just piping that into Sam's brain.
20:54
just like why why do I hear this in the background every time I talk email? It's so distracting. It's ridiculous. Um it's funny. Yeah.
21:06
I mean the other thing is like uh you know uh maybe the next AI product is just the the humanoid robot and you know Optimus just sits there. It's like C3PO.
21:16
I was watching Star Wars this weekend and and like you have a question you just ask C3PO.
21:20
The other thing you have your humanoid robot and you just ask the robot like the the opportunity with with uh the long-term opportunity with neurolink and AI is that you have a new uh sort of artificial internal monologue, right?
21:35
That basically is augmenting your existing thinking. Yeah.
21:39
And so you have to imagine that Johnny and Sam have thought about that potential scenario.
21:44
And so it's poss you could imagine they Apple's working with that company around uh uh sort of brain computer interfaces.
21:52
They could just buy that company for a billion dollars too, right? Yeah. There are a few players. Interesting.
22:00
Uh the deal is expected to complete this summer.
22:01
The takeover card embedded into my finger so I can just swipe.
22:06
You can actually do that.
22:06
Uh you can get an NFC chip in your under your skin embedded in like a little capsule.
22:14
our our people will it's pretty pretty talk to ramps people and biohacking we'll make it happen. Yeah. No, you can do it. You can do RFID.
22:19
I know someone who did it and has uh could like open open a door with it basically. Super helpful. Yeah. Super helpful. Yeah.
22:26
Way easier than a keyboard. Totally worth it.
22:30
And the really crazy thing is that uh the the the capsule is like metal and so your body like rejects that.
22:35
So they have to co coated in silicon.
22:36
But if the What about just replacing your fingernail with Yeah, I think there's probably other ways.
22:42
I'm not into the body modification stuff. It's not enough for me.
22:45
What about internal monologue?
22:47
You got I've spent a after that height mogging incident, you know, went went way down the rabbit hole around, you know, height surgery, try to go from 61 to 71 so I could Oh, wow. That would be you.
22:57
Um, but uh it would mean missing the show and do that.
23:00
Um, uh, you have an internal monologue? Yes.
23:04
I can you imagine not having one?
23:07
Like I cannot imagine not having an internal monologue to the degree that I think the whole story about no monologue people are like I think it's fake.
23:16
TVPN audience all has an internal monologue.
23:19
Is it just like an is it just like an IQ test basically or do you think the no monologue thing is real?
23:26
I think it's just entirely fake and just made up.
23:29
I'm very skeptical of it.
23:32
We should have If you don't have an internal monologue, come on. We'll break it down.
23:37
Anyway, anyways, 55 hardware engineers, software developers, and manufacturing experts, a team that will build what IV and Altman expect to be a family of devices.
23:47
The two executives have already been exploring some early ideas for about 2 years.
23:49
They said the pair expects the first device to be a truly novel type of product.
23:56
People have an appetite for something new, which is a reflection of a sort of unease with where we are currently.
23:59
I've said the question is just like Apple's really tested every possible screen size down to the watch is super tiny.
24:08
Then there's the phone, the iPad, the the the mean the YouTube machine. The YouTube machine.
24:15
Uh uh and then and then you put on the vision pro and the goggles and you can get an iMac screen.
24:21
And so like if the device has a screen, it's got to be something between one inch and 200 in.
24:28
and and quickly you fall in that continuum. Yeah.
24:31
I mean the question is are they building for GPT 40 or are they building for GPT six or five, right?
24:40
And and how much better do the models get and our screens I mean wasn't Sam also an investor in Humane. I think he was. I think so.
24:47
I mean I won't list the investors in Humane because some of them are friends and they wouldn't want to be associated necessarily anymore.
24:56
But I think Sam's been thinking about hardware in a in a in a bunch of different ways and wants to test both.
25:01
No screen, see if that works.
25:04
Maybe screen, maybe different different opportunities. I don't know.
25:08
Um Ivan Alman's first devices are slated to debut in 2026.
25:11
That's one thing months before ASI. Yeah. Yeah. Yeah. Right around the corner.
25:16
Um, one one thing I feel really confident in is that I believe they could sell a million of these devices almost immediately, right? Well, it depends.
25:27
I think I mean, yes, a million is like low enough that like it'll happen no matter what.
25:31
The the question is like if it's if it's additive and complimentary to my current tech stack, it's a no-brainer. Absolutely.
25:40
Everyone will will will get one.
25:42
if it's a replacement, it it really has to be also a good phone still.
25:45
And if they don't hit the basics of, okay, it it works with iMessage, then there's a question about like, am I really going to drop my phone or am I carrying two phones now if it's like a phone size device, right? Yeah.
25:59
Um whereas it was very easy to to add a Chromatic to my tech stack and it's like, yeah, I could probably get a Game Boy emulator on my phone, but it's just a cool device and I throw it in my backpack if I'm traveling. I have it.
26:12
Uh, same thing with like even the Vision Pro. It's like very additive.
26:16
It's not fully replacing your phone.
26:18
So, so VR headsets could be something like glasses, like the meta ray bands.
26:22
People add those to their stack.
26:24
It seems like it's easier to get someone to to add something than to replace something as important as their laptop or their phone, which they have like 10 years, 15 years of like daily driving use and expectations on where the apps are, where how things work.
26:38
It's going to be a lot harder to rip out the current flow. Yeah. I don't know.
26:42
When he left when he left Apple six years ago, IV started the firm Love from a collection a collective of designers and engineers.
26:48
Their staff includes veterans of Apple's hardware and software departments as well as friends of IV and other collaborators.
26:52
He then co-founded IO last year with Apple alumni Scott Cannon, Evans Hanky, and Tan Tang.
26:59
Hanky was IV's successor at Apple and remained at the company until 2023, while Tan led iPhone and Apple Watch product design until 2024. Absolutely. Stacked roster.
27:09
Canon worked at Apple before co-creating the once popular email app Mailbox, which was acquired by Dropbox. Oh, I had no idea.
27:19
Did you ever use Mailbox? No. Fantastic app.
27:21
Uh 2013 comes out and it and it invents the the swipe to archive, swipe to snooze um uh uh paradigm and really makes it easy to start doing inbox zero.
27:35
And that's where I learned inbox zero because when you signed up for mailbox, there was this big long wait list.
27:38
It took you a long time to get through.
27:39
It was very like it was a hot product at the time.
27:42
And one of the first things it would do is just say, "Hey, you have 5,000 emails. Let's archive them all."
27:49
Because when you archive an email, it just leaves your inbox.
27:50
You can still search for it. It still exists. It's not deleting it.
27:53
So, let's archive it all.
27:55
And then from now on, you should triage your email by swiping.
27:57
And it was it was amazing. It was a great product.
27:59
And then mailbox bought it or Dropbox bought it and I think they shut it down.
28:02
It was one of those Jurro sushi products, right? Totally.
28:06
It was like really great design and everyone loved it for like the year that it was viable and then I don't know what happened with it.
28:12
Um, at IO, the group set out to to develop, engineer, and manufacture a collection of products for an era of of artificial general intelligence, the point at which technology achieves humanlike cognitive abilities.
28:23
The team will now continue that mission at OpenAI, becoming a threat to the very devices the designers help create.
28:28
That adds to the challenges.
28:30
Apple, which has fallen behind in Silicon Valley peers in artificial intelligence.
28:33
The the company's AI platform released last year lacks the capabilities of rival systems and relies in part on OpenAI's chat GPT chatbot to fill in the gaps.
28:41
Still, Ivan Altman don't see the iPhone disappearing anytime soon.
28:45
In the same way that the smartphone didn't make the laptop go away, I don't think our first thing is going to make the smartphone go away.
28:52
Okay, little getting a little bit more detail here.
28:54
It's a totally new kind of thing.
28:56
I'm really excited for this.
28:56
Um, the phone as it as it is as it currently is is a remarkable general purpose device.
29:05
I've said adding that people will connect with AI in very new ways.
29:09
We got to get uh I'm so excited to to understand what the interface for this first device is going to look like because one thing stood out uh Nikita was posting earlier that if you're onboarding flow requires any sound. Yeah.
29:23
you just see an immediate massive reduction in usage because a lot of people use their phones.
29:27
What's an example of of of an onboarding flow that requires sound?
29:31
Uh some apps will play a video, but I'm talking about this in the context of like a voice interface with the app.
29:38
So the issue is people use their devices in bed.
29:40
They don't want to be loud. Got it.
29:42
Um and they don't want to be loud.
29:45
So if it's if if it if the experience requires sound, you're cooked. Yeah.
29:50
He Nikita says the majority of new app installs take place in bed or on the toilet.
29:54
If your app requires sound on or voice to sign up, you're voluntar you've voluntarily accepted a 60% tax on your growth.
30:00
And so that's you know if your voice uh voice AI has progressed so much, right?
30:10
We've seen examples of sesame super low latency. Yeah.
30:12
uh you could imagine a hardware device that's purely allowing you to interact with with AI like it's a human, right?
30:20
What if the device is a po a response to all the all the fears around AirPods and so it's a set of wired headphones into a Walkmanl like device and then you can talk to the AI without needing to use wireless headphones.
30:36
That could be really trendy.
30:38
They could go back to the original iPod video of the silhouettes dancing with the with the headphones on.
30:42
So TJ has a great post here.
30:44
Open AAI got the best hardware and design team in history for six billion or frame differently for the same amount Amazon burns annually on Alexa and devices. Yeah. Yeah.
30:54
No, I I I really think that this is not a crazy overpriced acquisition.
31:00
It seems totally reasonable to take 2% of your equity and say, "Hey, let's take a shot at owning the next computing platform because this is such a seismic shift that that there's totally a world where they go and solve glasses or watches or headphones or some sort of other mobile device that you carry."
31:18
Uh, and maybe it's additive.
31:21
We should get some of the device uh the device entrepreneurs on the show. Obvious tweets.
31:28
I'm sure he's launching in 71 days. 71 days.
31:31
He just posted he just posted a minute ago. Too many cooks.
31:36
Too many cooks in the kitchen.
31:39
Which is funny because is that a play on Tim Cook? Yeah, maybe.
31:42
You know, also there's that Adult Swim uh video. Too many cooks. I don't know.
31:47
Too many cooks in the kitchen.
31:48
Obby I'm very excited for his launch. Yeah.
31:51
Yeah, we should give him a call.
31:54
Uh, and it would be even if even if OpenAI and the IO team just completely come out with their own version version of Friend, it's still a cultural victory for for Obby because he he got the form factor right.
32:06
But and he can probably sell the Oracle for a billion dollars or Amazon or something. Yeah.
32:12
I mean, the Alexa friend, that'd be good. The Alexa friend. Yeah.
32:15
OpenAI also shook up management ranks this month with the San Francisco based company appointing Instacart chief Fijiimo as the CEO of applications.
32:24
She reports directly to Altman allowing him to focus on the broader strategy.
32:28
Hanky who will become an OpenAI employee along with Tan and Canon said that Chat GPT's debut prompted a realization that hardware technology would have to change.
32:37
A number of us looked at each other and said this is probably the most incredible technology of our career.
32:40
So agree but how will it change? Tell us something. Just one thing. Just one thing. Just a crumb, sir. Just give us a weight. How much will it weigh? How big will it be? That would be fun.
32:53
Is it bigger than a bread box?
32:54
Can we do 20 questions with one of these?
32:55
20 questions with That's what we'll do with Sam.
32:58
20 made out of metal or plastic or is it organic? Can I eat it? Is it a robot?
33:07
Does it This is This is interesting.
33:07
So, so Sam says, "We are obviously still in the terminal phase of AI interactions.
33:14
We have not yet figured out what the equivalent of the graphical user interface is going to be, but we will. Mhm.
33:20
Is that saying that there's a should What's interesting is like the the graphical user app, right?
33:25
The graphical user interface didn't change with mobile.
33:31
The thing that that the iPhone did well was multi-touch.
33:33
And so being able to like pinch to zoom and all of that was like well no I mean I mean obviously the breakaway app was when you could just drink the beer drink the beer the beer app was the was really the win.
33:45
I mean that the iPhone wouldn't I mean we've already had that with AI that's the Harry Potter Balenciaga that that is the iPhone beer app of of AI.
33:51
I think that's a good take.
33:55
Uh Love from has a number of former Apple designers who helped create the look of the Mac and iPhone operating systems that haven't changed in a decade since they did Somebody go check on John Gandrea.
34:05
Oh, we never finished that article.
34:09
I really want to finish it.
34:11
Uh they they could help redesign OpenAI's app for new generations of consumer consumers.
34:15
Uh Love from will continue its existing relationships like Ferrari and and Airbnb, but it won't take on new clients.
34:23
Their new hardware team within OpenAI will be overseen by Peter Wlander, who will report to Altman as product vice president.
34:29
They should just buy Ferrari and Airbnb.
34:31
Can you imagine if OpenAI own this is the new device?
34:36
If you want to interact with chat GPT, we recommend getting in SF9 GTB SF90 Chat GPT integrated. Yeah.
34:46
I mean, you're already paying $2,000 a month for ChatGpt Pro Plus.
34:51
Why not put it up to $4,000 and get a lease for a SF90? Yeah.
34:55
Ferrari will finance it for you. Exactly. Yeah. Totally doable.
34:59
For the price of a chat GPT subscription, you could easily be driving a hypercar. Totally. Totally. Yeah.
35:06
Everybody talks about the $20,000 a month, you know, chat GPT subscription.
35:10
That should come with a Ferrari. Infamous. Yeah.
35:12
It should come with an SP3, right?
35:13
It should just absolutely.
35:15
It's not humanoids, folks.
35:17
It's AI will be embodied by the legacy of Enzo Ferrar.
35:21
Think about how low churn would be if you were like, in order to use AI really effectively, I got to be revving that V12.
35:27
a guy who opened it up on the open road.
35:30
The device needs to be vibrating from the naturally aspirated V12 cutting canyons.
35:36
Well, I mean there there was uh didn't Elon say something yesterday that uh it wasn't off the table that that uh and X could merge. Yeah. Yeah. Yeah.
35:45
Luke Metro was was predicting that a year ago. I'd finally buy a Tesla. Yeah.
35:49
So I could be terminally online behind the wheel. Yeah. I like those.
35:55
Roll in Ferrari for sure. You heard it here first. Breaking news.
35:57
I've said Chachi PT was initially put on his radar by one of his twin sons, Charlie. Uh oh, twin dad.
36:03
Let's hear for twin dads. Love that.
36:06
Uh and he immediately knew he had to meet Altman after using it.
36:08
The new team at OpenAI will work at IO's existing workspace in Jackson Square, a neighborhood in San Francisco and OpenAI's current offices, so they're not far apart.
36:16
I have felt that my most important and useful work is ahead, I've said, adding that he has been training for this moment.
36:22
He compares the experience to Apple in the late 90s to early 2000s before the iPod and iPhone.
36:29
I'm really I'm just really really grateful we all found each other.
36:31
Ivan Altman wouldn't elaborate on what hardware products they are working on, but they will be entering a market in its infancy.
36:39
Meta Platforms, the owner of Facebook and Instagram, is perhaps the most notable maker of AI devices.
36:43
It sells the popular Ray-B band smart glasses, which Apple uh which Google's now competing with.
36:49
They launched two different uh devices.
36:51
Enreal, these VR glasses that uh it's it's not full virtual reality.
36:57
It just projects a really big screen in front of you.
36:58
I've been wanting to pick one of these up because my main use case for the Apple Vision Pro is just watch a movie on a plane.
37:05
And these are like cheaper, lighter, easier to use, and they just put the big screen in front of you. Very cool.
37:11
Um I want to get some of those.
37:13
And we should probably try and get the CEO on the show.
37:14
And then uh and then they also launched um like like like smart sunglasses that don't have a display but you can talk to Gemini which also very cool.
37:23
I've took shots at Humane and Rabbit saying those were very poor products.
37:32
And says there's a quote brutal.
37:34
There has been absence of new ways of thinking expressed in products.
37:39
I is that the problem though?
37:41
Like those were definitely new ways of thinking expressed in products.
37:45
They just weren't good new ways of thinking.
37:47
They were definitely new though.
37:49
Not every new way is going to be a good way. Yeah.
37:50
I also think that I still feel like the humane AI pin built for children could have the Rabbit R1 specifically too.
38:00
Like it it would be so easy to sell that to your your your your four-year-old, your 5-year-old is going to start demanding an iPad.
38:10
instead get them this and it's the ner it's nerfed iPad and it's just educational.
38:15
It's a machine for your child's curiosity.
38:17
They can ask it any question any questions.
38:19
They can say tell me a story about play Skinner box games and they can't play Candy Crush on it or whatever. Yeah. Um Yeah.
38:25
No, there's definitely something they'll figure out a way to put Candy Crush on there.
38:30
They will some of these whale kids spending 20 grand a day on Candy Crush.
38:34
But but it's a it's a vending Johnny's like this was not my intention. Yeah.
38:39
Yeah, he just creates the most addicting advice device possible.
38:40
Um, they're calling it the new cigarettes, folks.
38:44
The um Ivan Altman wouldn't elaborate on what they're working on.
38:49
Um, Tan, who is a central uh who is central to developing every version of the iPhone within Apple's hardware engineering department, said the new team isn't tied to a legacy and will have an opportunity to rethink this space.
39:00
It'll be worth the wait, Alman said. It's crazy.
39:02
It's a crazy ambitious thing to make.
39:05
Not much detail here, but thanks for the post. I am excited. I hope it's huge. I hope it's a hit. Really, really big.
39:16
I want it to be as big as a newspaper.
39:17
I think part of part of really loving big tech is wanting more companies to become big tech. Yeah. Or high technology.
39:25
And OpenAI has clear potential to get into that group.
39:30
And maybe they should launch out a newspaper.
39:34
This is kind of hardware.
39:35
Just print out deep research reports. Yeah. Just mail it to me.
39:37
Put it drop it off on people discount paper as as a, you know, a potential hardware form factor. Yeah, it is. It is.
39:44
Anyway, uh Google IO reactions. Let's move on.
39:49
Um, everyone on the Google team made the exact same hand gesture.
39:52
Apparently, they were all trained. Tulsi in there. Tulsi's in there.
39:55
She's been She was on the show yesterday.
39:57
I guess Logan didn't make that. No.
39:58
Um, maybe he had some wild hand gestures. Says goo gang sign. gang sang.
40:04
Uh, are you familiar with the Merkel route?
40:06
This one, Angala Merkel would always hold her hands like this. It was pretty sick. Similar to Andrew Tate.
40:12
Is that where he got it from? Wait, Tate does it? Oh, I didn't know that. That's hilarious.
40:16
Uh, I think I think his his infamous Oh, he does he does like this. This tents his fingers.
40:22
Tenting your fingers is very like pensive and thinking.
40:26
I'm thinking, "Oh, yes, I'm tenting my fingers."
40:27
It's not very golden retriever coded.
40:28
You You never see a golden retriever do that. No, definitely not.
40:33
Uh um there were some other news before there were founders waking up that to realize that their startup is now a bullet point in Google's keynote. Yeah.
40:39
So couple couple things here.
40:41
They launched a product to generate product designs. Yep. It was interesting.
40:47
They launched a dogee competitor which is like a tryon product.
40:50
Uh they obviously launched jewels coding agent.
40:53
they launched the eyewear, you know, basically whether you're Sesame, do y uh even Figma, right?
41:00
Figma has generative AI um tools. Yeah.
41:04
I I feel like but but again, I don't think that founders I don't think like the Dogee founders are seeing that and thinking, "Oh, I'm suddenly worried." Right.
41:17
Because the idea that Google is going to dominate in fashion, which is so culture driven, is seems really I I don't I don't see it happening.
41:26
I can see them having Yeah.
41:28
a version of the product in Google Google shopping that enables tryons.
41:32
I don't think it's going to be this like social cultural thing that that Doge is clearly tapped into. Yeah.
41:39
I wonder how many startups Google has actually killed in history.
41:43
It doesn't feel like that many.
41:43
The famous example, of course, is like Microsoft really did uh kind of kill off uh Slack's growth with Teams. Yeah. Right.
41:51
Um, and Slack kind of like, you know, reached its peak and kind of flatlined after uh after Microsoft Teams took off because Microsoft had such a distribution advantage.
42:03
I feel like there's two different types of startups.
42:06
Teams took off because defense tech got cool and all the defense tech had to bear comp.
42:10
Um, but but seriously, it's like I think that if you if you saw a Google feature and you were like, I'm going to start a startup on that and you're basically copying Google because you're like, Google's slow, then yes, this is real.
42:25
Like to all the companies that were like putting up fake Notebook LM apps that weren't really like raising money and kind of just like cash grabbing on top of like wrapping stuff that's already been created and pioneered by Google.
42:38
If Google drops a a notebook LM app and they like wipe you, you can't really be that upset.
42:46
Like you're always in a cash grab.
42:47
But if you have built like a massive design tool like Figma and it has so many different customers, so many different things and then they come out with like oh something that allows you to do like a little bit of design work like it doesn't feel like as nearly as existential.
43:01
Um and so I feel like this narrative is always like oh Google just like you know completely destroyed this stuff.
43:08
And it really depends on like is is this is this truly aligned with Google's mission that they're going to just keep pouring relentlessly at and they're going to make this great and it's and it's completely on their road map.
43:19
Um and were you always building a feature or were you building something that can that can grow into its own behemoth?
43:25
And if if you're on if you're on the right track, you're probably fine.
43:28
Um anyway, um what's next?
43:28
Well, speaking of behemoth, the US formerly accepts luxury jet from Qatar for Trump. No way.
43:40
Uh, and it will possibly be used as Air Force One.
43:44
And Yahoo is reporting that turns out Trump's Qatar private jet wasn't even a gift.
43:56
It wasn't, despite the insistence that the plane was a gift.
43:59
What was it then if it's not a gift?
44:02
uh payment for a senior White House official told CNN that Trump tasked Steve Witco, the president's special envoy to the Middle East, with tracking down a replacement for Air Force One after Trump learned that Boeing would not have new jets ready for another two years.
44:15
Witkoff ended up leading initial conversations with the Qatari government.
44:19
They have them in stock, right? Yeah.
44:20
Boeing provided the Pentagon with a list of other clients who might be able to help with America's search for a new plane.
44:27
One of those sources says that Qatar was included on that list of clients.
44:29
And remember, we saw that chart, right? Yeah. Yeah.
44:32
Uh there's not that many people in the world that actually have these jets set up in this way.
44:37
Um there were also discussions about leasing the planes.
44:42
So again, uh when we saw the initial uh headlines, uh we thought that uh we imagined that it it was possible that the the full truth was um not uh not really revealed yet.
44:52
And apparently legal negotiations over the plane's transfer are still ongoing.
44:57
Uh, and it's unclear how the plane went from being a potential purchase to a $400 million gift.
45:07
So anyway, where should we go next? Probably ramp. Time is money. Save both.
45:14
Easy to use corporate cards, bill payments, accounting, and a whole lot more all in one place.
45:17
We already talked about Figma, but let's tell you about Figma as well.
45:20
Think bigger, build faster.
45:22
Figma builds helps design and development teams build great products together. Get started for free. the backbone of TBPN. Yeah.
45:29
If you like some of the merch that you've seen, that's where we're making it. Oh, yeah. Can do it all.
45:36
Let's go to this Gary Tan post. This is good.
45:38
Actually, talking about the cycle never ends.
45:39
Every month, a new world's most powerful model takes the crown.
45:44
We applaud, compare, and then wait for the next.
45:46
OpenAI introduces the world most powerful model, then Anthropic, then Gemini, then Grock.
45:51
There's a little bit of fracturing now in the sense that I haven't played with Gemini's LLM enough to displace my OpenAI workflow, but the generative video like VO3 is truly way way better than Sora right now.
46:06
And so I you compared the two I I same prompt in both and it was brutal.
46:12
Not there was not even a question which one you would go with. Yeah.
46:17
One the Sora output made me kind of it was it was a lot of hallucinations tell story. It didn't have movement.
46:28
It was definitely like like a lot of gyrations and just hallucinations.
46:32
Now that's not to say that the V3 V3 had a lot of crazy hallucinations, but they were not as disturbing or more like psychedelic.
46:44
like we we we we generated this one video of uh of an Audi R8 with a TBPN livery pulling out of our new studio and it peels out.
46:52
It's very cool, but then all of a sudden it's driving backwards, which is very odd and jarring.
46:56
Uh I did another one with a a a Rolls-Royce Phantom pulling into the studio, the TBPN GT3 Cup Car livery foresight.
47:03
And that foreshadowing and that one when when it pulled in, did you notice this?
47:11
It had a wing on the back and it pulls in and then all of a sudden the car morphs and you're looking at the front of the car again.
47:16
And so it kind of hallucinated and transformed.
47:18
It did that a couple times where us blasting through the Hollywood sign.
47:23
It we there's a Hollywood sign in the background.
47:25
We bust through one Hollywood sign, then there's a third Hollywood sign.
47:28
Like it is still hallucinating.
47:30
It doesn't seem it doesn't seem grounded in reality in any way, but just visually the the the fidelity of the visuals is incredible.
47:39
and the ability to to include three or four different concepts and then it weaves all of that into an eight into an 8-second clip.
47:45
It's very very cool and the sound is obviously amazing.
47:47
And so I have been begging uh Google representatives for more V3 credits because it's a lot of fun.
47:52
Um but yeah, we love we we we love a we love a an AI race.
47:59
Well, we have some kind of meta news.
48:02
Somebody is tr doing live translation of TBPN in Dutch right now.
48:05
Do you want to try playing it, Ben?
48:11
Oh, we'll see if this works.
48:16
And we have Anthony Pompiano joining at uh at 12:30 now.
48:18
So, uh in the meantime, while we're pulling that up, let's tell you about Vanta.
48:24
Automate compliance, manage risk, improve trust continuously.
48:28
Vanta's trust management platform takes the work out of your the manual work out of your security and compliance process and replaces it with continuous automation.
48:34
Whether you're pursuing your first framework or managing a complex program.
48:39
Um, I need Pomp's email for this invite.
48:42
Let's also tell you about Linear.
48:43
Linear is a purpose-built tool for planning and building products.
48:45
Meet the system for modern software development.
48:48
Streamline issues, projects, and product road maps.
48:51
Linear launched uh their agents product into GA today, which is very cool.
48:57
And I want to ask Scott about it because uh they have been working on some serious integrations there which would be cool.
49:05
Um we uh Wow, they're translating. That's pretty.
49:15
It's going to get really gonna get really freaky once they get to this point in the stream and they're translating us and it becomes very meta.
49:24
Oh wow, they're actually talking about Vanta now. This is incredible.
49:27
Fantas for this is really impressive.
49:30
Okay, anyway, let's move on.
49:32
Enjoy that if you're looking to for TVPN in Dutch.
49:35
Um, Chubby shares, "This is freaking insane.
49:39
The holy grail of personalized AI for daily usage.
49:41
If this demo turns out to deliver what it promises, will be much further ahead in 2025 than I would have thought."
49:46
So, this is a video of a guy who uses a Android phone to troubleshoot a bike that he's trying to fix.
49:52
And he asks the AI to pull up the actual manual or uh yeah, manual for the bike.
49:58
And it's scrolling the PDF and then it says, "Hey, go find a YouTube video and it opens the YouTube app."
50:05
So, it's very much working at the OS level to use all of the features that are available on any Android phone.
50:12
And that seems like what we want from Apple intelligence that we're not getting yet.
50:17
You can't tell Siri to go open the notes app and search there or take a new note.
50:24
Everything has to be puppeteered at like the API level through these specific integrations instead of just using the whole the whole phone.
50:30
And so uh it's a very cool it's a very cool demo.
50:34
There are a bunch more uh pieces of news.
50:36
Uh Sam from Futurenomics says Whimo does more rides than Lyft.
50:41
In San Francisco, there are only 300 Whimos.
50:43
Lyft has 400 uh 45,000 drivers.
50:46
Every Whimo is doing more than more rides than 150 human drivers.
50:51
Now, there's a community note on this.
50:52
I don't know how true it is, but there were some staggering Whimo numbers.
50:54
I think they crossed 10 million rides or something.
50:56
There's this comment right underneath it that I think is probably why it's getting noted, which is that uh this guy William says, "That number feels implausibly high, though.
51:05
I wonder if that lift number includes surrounding area drivers. Occasionally do ride.
51:10
There's no way 45 lift has 45,000 drivers in San Francisco.
51:13
The streets would be packed.
51:16
It's it's what a city of like 7 million people or something like that.
51:20
Like a lot of pink mustaches and and reminder like Uber's what a 100 times bigger than so you four million Uber drivers for 7 million people or something like that. Yeah, it makes no sense.
51:30
Anyway, uh Olivia Moore over at Andrea and Horowitz says uh although she lift is flat on the last year on the last 12 months.
51:41
So, uh it's up actually up 1 cent over the past year.
51:45
So, anyway, we need a poly market on the current best AI model.
51:52
I want to see where that's trending.
51:55
Um because Google obviously released a bunch of models and Google was in the in the in the leading contender for top model and I want to see where the poly market shaking end of May Google's sitting at 91% 91% for Google 40% for December 31st end of the year. Yeah.
52:13
So Olivia Moore here says Google had a bunch of great AI launches today but they're so hard for consumers to access.
52:19
I completely agree with this.
52:20
I get stuck all the time in the wrong flow in the wrong app. even accessing V3.
52:24
I wound up in Flow, which is their TV product or something, and it was just showing me AI stuff.
52:29
And then there's like I can get into the Dev Studio, but then I'm have to set up an API key.
52:35
And then I finally have the the Gemini app, but the Gemini app wasn't wasn't upgraded.
52:39
And so I had to go through a flow on on Chrome on desktop to get that.
52:44
And I get that I don't have an Android phone, but I feel like I'm being punished too much. Yeah.
52:48
Um, the bull case there is that that kind of stuff is actually easier to figure out than being on the technological edge and leading in benchmarks and all these other things and having massive Yes. Yes.
53:01
But they should be able to do both.
53:02
They have so many No, I totally agree.
53:04
I'm just saying it's it's it's uh, you know, talking to Tulsi and Logan yesterday, they clearly are credibly locked in, aware of aware of the the push back. Yep.
53:12
So, uh, she goes on to say, "It feels like every debut comes with a new subscription, which you need a specific account type to access and have to hunt down the playground for.
53:20
I know AI products are expensive to run, but with Google's distribution and revenue streams, I don't get why they don't offer some of some more of this for free." That was my question.
53:28
Uh, or at least make it easier to find.
53:32
Notebook Notebook LM feels like their last big viral launch, which had no payw wall at various times in the last three months.
53:37
I've been subscribed to Gemini Advanced, Google One AI Pro, and now AI Ultra.
53:42
The madness needs to end. I agree.
53:43
Although I'm now an AI Ultra subscriber, so maybe they got me.
53:45
Maybe it's all part of the plan. 4D Chess.
53:47
Um, yeah, I'm interested to track the ramp data on Gemini over the next couple months.
53:54
See if See if it's See if it's ramping up.
53:56
Um, anyway, uh, everyone's going to have to pay sales tax on this. Get on Numero. Go to numeralhq. com.
54:00
Put your sales tax on autopilot.
54:02
Spend less than five minutes per month on sales tax compliance.
54:05
Uh, Wes Roth says, "Google IO 2025 was an absolute AI overload and everyone is still picking up their jaws off the floor." Here's the TLDDR.
54:12
Gemini gets an agent mode.
54:14
It's so powerful it borders on executive assistant.
54:16
So, excited to check that out. Uh, post from Addie.
54:21
Gemini is the kid that got bullied in high school and is now jacked a millionaire and practices gratitude. It's very accurate. Yeah.
54:27
It's very accurate. Yeah. I mean, they just like I mean, that's the nature of these companies that that are on the annual releases is that like you you're like they're behind for 11 months out of the year and then they're
54:37
state-of-the-art for the big release and then of course everyone just like yeah, I'm just like feeling extremely grateful that you know we own you know 80% of you know what you know as the internet and you know we can leverage this into creating you know really great AI. Yep. Yep.
54:53
Uh Theo says, "Google buying DeepMind in 2014 for $400 to $650 million, might have been the greatest acquisition of all time." Completely agree.
55:00
It's definitely up there.
55:03
Definitely in the conversation.
55:04
Definitely in the conversation.
55:05
Um Signal says, "Google's edge in local data, especially now with LLM, is absurdly underrated.
55:09
Nobody touches the depth, historic history, and granularity they have on local businesses, landmarks, real-time conditions.
55:16
Every other mapping service is just a sad knockoff because Google's been building this mode for decades." Overnight success.
55:20
Also, I mean the YouTube data is super valuable.
55:23
You have to imagine that their edge and VO is because of the YouTube data because I mean like it's the biggest repository of video content new stream of video constantly.
55:35
Constantly constantly right. Yeah.
55:37
And so that's probably why uh there's so much good uh stuff coming out of V3.
55:41
Ben Hilac said showed the first ever ad produced.
55:45
Welcome to the token tavern.
55:47
It's in the tab if you guys want to try and pull that up.
55:49
Uh Justine Moore Venture Twins says, "Man singing a song about being capitalist."
55:53
We should play both of these.
55:55
Uh good Alexander uh contextualizes it by saying, "You can own Google at 17x earnings or short and short Costco at 50x earnings.
56:02
Long AGI short rotisserie chicken is actually value investing."
56:09
People love People love Costco.
56:09
Uh and then Archie Sangupta says, "Got access to Google Diffusion.
56:14
Holy 99 tokens per second.
56:18
I made a calendar in three seconds. Yeah.
56:19
I mean this is this is what people are talking about in terms of being able to generate entire applications, right?
56:26
And this is the risk to existing software applications that don't have network effects are really sort of thin wrappers around AWS and you know these sort of simple applications.
56:41
I mean, I still think that um this will be one of those things, you know, people uh people can make their own food.
56:50
They still order out from restaurants, right?
56:54
So, here, let's some of these with audio if we can. The token tavern. It looks so good.
57:01
I don't think I'm supposed to be here.
57:04
Well, partner, that's why I use raindrop, so I can find bugs like you before they bother my users.
57:11
Let's get the hell out of my town. Wow. Ben cooked. Cooked. Ben cooked.
57:19
Although Although though I know that those graphics came from After Effects, that overlay for sure.
57:26
After Effects not generated produced with V3. Yeah. Yeah, it uses V3.
57:32
This is the most impressive shot in the whole thing though. Look at this. Perfect lip sync. The teeth look great.
57:38
the voice, the texture, but get an alert when your AI product fails.
57:43
That's motion graphics, baby. Truth Zone. Truth Zone.
57:45
Ben Cook, congrats to Ben. That's a very cool ad. Very cool.
57:50
What about the other one?
57:52
This one I I basically watched this and I said I had to watch this.
57:54
I'm going to make John watch this.
57:56
It is pretty rough on the ears. Yeah. Okay, let's lock in. Check it out.
58:01
Uh, in the meantime, let me tell you about public investing for those who take it seriously.
58:04
They got multiasset investing industryleading yields. Trusted by millions. Public. com. Public. com. Get on there. Also, how'd you sleep?
58:18
You know how I slept, John. I had to come back.
58:20
You've You've had a couple good nights.
58:22
I had to show you who's boss. I think I did poorly. 90. Not bad. 97. Oh, I got you. I'm back in the game. I'm back in the game.
58:30
Uh, autopilot made 22 different 5 ultra.
58:34
Autopilot made 22 adjustments to the temperature of the bed to get my REM sleep up 8%. Just last night. Just last night. 22. That's actually a lot. Okay.
58:44
Anyway, let's play this next video.
58:47
Thank you to Eight Sleep. [Music] All right. Turn it off. Turn it off. Turn it off. This is so bad. So bad. That's wrong. That's just wrong.
59:10
Let's just end the stream.
59:13
Yeah, just just turn it off. That's enough. That's enough for today.
59:16
Honestly, let's take the rest of the week off, too. Just hang it up. That's so bad. Yeah.
59:21
Yeah, I was looking at that viral that viral VO video of and the prompt was just like standup comic telling a joke and everyone was like, "Hey, this is actually funny."
59:29
Uh, it's like I went to I went to the zoo. They only had a dog. It was a zoo.
59:35
Funny, but like that joke is not created from AI.
59:39
It's literally on Reddit.
59:41
It's literally at the top of like Reddit rkes.
59:43
It's like we still we're still training on the same data.
59:47
Okay, somebody somebody take that video of somebody take that video of the the VC singing that song, post it on Blue Sky and said and just say like, "Hey, I know you guys haven't been on X in a while and but like here's what's going on."
1:00:00
And given that they're generally decelerationists, I think they might actually Yeah, they're not going to be happy with that. Bad.
1:00:07
They hate to see a young venture capitalist singing. Very bad. Just just go to adquick. com. Buy a billboard.
1:00:17
Out of home advertising made easy and measurable.
1:00:20
Say goodbye to the headaches of out of home advertising.
1:00:22
Only Adqubines technology, out of home expertise, and data to enable efficient, seamless ad buying across the globe. Thank you. That was rough. Yeah.
1:00:29
Anyway, we have three minutes.
1:00:32
Let's just do a little quick timeline.
1:00:34
We'll get to the Ben Thompson later.
1:00:36
Um, but I wanted to cover this article in the Wall Street Journal.
1:00:40
Europe's firms fall far behind in the race to create big tech.
1:00:44
Lack of innovation holds economies back, weighing on the continent's future.
1:00:48
Uh Berlin, the world's technology revolution, is leaving Europe behind.
1:00:55
Europe lacks any homegrown alternatives to the likes of Google, Amazon, or Meta.
1:01:00
Apple's market value is bigger than the entire German stock market.
1:01:02
The continent's inability to create more big technology firms is seen as one of its biggest challenges and is a major reason why the economies are stagnating.
1:01:10
And so if you look at their chart uh and they're looking at number of unicorns by region in the United States there are 90 unicorns with a combined value of 2. 35 trillion.
1:01:20
SpaceX and OpenAI of course the biggest uh in the EU there's 107 unicorns.
1:01:25
So actually more companies but their combined value is 330 billion.
1:01:30
Basically an order of magnitude less value.
1:01:32
And they include Stripe in Europe, which is not fair in my opinion because Stripe is a YC company that's very San Francisco coded.
1:01:39
Uh they have their conference.
1:01:41
They do in San Francisco.
1:01:43
In San Francisco, but I've been thinking more about European venture capital because a lot of people have been hating on it, thinking that it's impossible to make money there.
1:01:50
This is kind of consensus.
1:01:50
All the smartest venture capitalists think that it's just impossible.
1:01:53
But there is something amazing about that type of radical risk takingaking in such a challenging environment.
1:02:00
Totally like and if you just think about like the act of launching a VC fund in Europe, it goes back to the early spirit of Silicon Valley. Totally.
1:02:08
It's totally like a huge this idea of giving money to a company, giving millions of dollars to a company that has no cash flows that's maybe entering a market that doesn't even exist yet. Yeah.
1:02:17
And um it's a million to one investing in in European companies.
1:02:22
It's a million to one that you even make a dollar of revenue.
1:02:25
Totally potentially a billion to one.
1:02:26
You know, people talk about like one miracle, one miracle companies, two miracle companies. Yeah.
1:02:30
In Europe, you need to be a 10 miracle company. Totally. To do anything.
1:02:34
And so, if a VC can go there and even return even just a 1x on their fund, I think that they should be at the top of the minus list because it is such a remarkable achievement.
1:02:45
It's so running the Midas list these days. I don't know. Postrad Conrad.
1:02:49
But it truly it like given the given the expectations if you can if you can even get out of there with returning 50% of your fund. Yeah.
1:02:56
You should be you should be you should be remembered as one of the greatest venture capitalist to ever do it.
1:03:01
I think it would be it would be it would be a remarkable accomplishment.
1:03:03
You need a reality distortion field. Yeah.
1:03:05
To just to just go and deploy in Europe. It's remarkable. Some level of hubris. Yeah. Massive.
1:03:12
Anyway, anyways, we have Eric Prince in the studio.
1:03:14
Why don't we bring him in? Eric, how you doing? There he is. What's happening?
1:03:21
Welcome to the stream, gentlemen. How are you? We're great. How are you? Great to have you. How are you? Thank you.
1:03:24
Um why don't uh I don't know if you need any introduction, but uh why don't you give an introduction uh for those that that may be unfamiliar.
1:03:39
uh born in Michigan uh served in the Navy for a while and built a business called Blackwater which was a defense services firm, a private military contractor if you will.
1:03:48
I sold that in 210 and I've done some things in um resource investing upstream in mining and uh exploration and I still uh I help governments from time to time with large intractable problems and uh started a company just over three years ago called Unplugged which is a standalone independent phone outside of the Google and Apple universe.
1:04:15
It's really kind of an an answer to surveillance capitalism. Mhm.
1:04:18
What was the first dollar of revenue for Blackwater?
1:04:23
Like how do you get that business off the ground?
1:04:27
Uh so SEAL teams, special operations units, because of the gun culture in America have gone to private facilities really since the 70s.
1:04:35
Some, you know, some champion shooter that's running a shooting school somewhere.
1:04:40
And no one had done that on an industrial scale.
1:04:45
And so I knew coming out of the SEAL teams, I knew what the teams needed.
1:04:46
and I built the largest facility of its kind in the world within an hour drive of Norfick Naval Base and all the SEAL teams and military concentration that's there.
1:04:58
And actually our first customer was from the West Coast, a West Coast SEAL team, but um yeah, we we built ourselves into becoming a indispensable part of the trainup of military units.
1:05:12
And then when uh Coline High School happened, that terrible high school shooting, you had dozens of police departments that responded and didn't do very well because it took them hours to clear the building and people bled out that should have been cared for.
1:05:24
And so we ended up building a big mockup called Are You Ready High School where we could train thousands of SWAT officers how to respond to those kind of disasters in process with all the sounds and lights and role players and screaming and um and that was very effective.
1:05:42
And then with the USS Cole Navy ship was blown up in Yemen by a suicide boat.
1:05:49
The Navy realized that well the the sailors that were guarding that ship were holding largely unloaded weapons that they never fired before because the Navy thought it was too dangerous to train with live guns.
1:05:59
And so we ended up training a 100,000 sailors.
1:06:01
That was really our first big defense contract. Interesting.
1:06:05
And then and then when 911 happens obviously the demand signal for security and training and aviation and uh you know operating in difficult places was our specialty.
1:06:17
How do those contracts actually work?
1:06:18
We're familiar with the defense tech pipeline.
1:06:20
A lot of defense tech startups right now maybe they're drilled they're building counter UAS drone systems and the typical path is they get some SBIR money then they move to program of record they start scaling up manufacturing but for what you're doing I imagine that the contracts are structured differently.
1:06:38
How do you project revenue?
1:06:41
How do you think about ongoing relationships with the DoD and making sure that the customer is satisfied for a long time?
1:06:51
Uh it's a much shorter flash to bang uh in terms of uh demand signal.
1:06:54
A customer would come and say I need this kind of airlift in this part of the world uh move this many tons.
1:07:01
of the world uh move this many tons. um need to be able to fly on night vision goggles and operate under this level of uh of FAA or not scrutiny and we would bid it about 98% of our revenue was competitively bid and I look at at the same time I was building Blackwater I
1:07:22
took over my dad's original business which made diecast machines a big you know heavy machine tool business and I took that through a lean transformation process really based on the Toyota production system and you lowering the cycle time of large quantities of steel and other work in process. And so as I'm
1:07:39
And so as I'm laying out Blackwater, I kind of thought about it like a like a factory.
1:07:44
And what does a military do?
1:07:47
It recruits, vets, equipss, trains, deploys, and supports people to do a job at a difficult place. That's what we did.
1:07:54
And because we were a training facility to start with, we could back into the recruiting and the vetting.
1:08:01
We had our own in-house uh doctors and psychologists for all the psych screening tests, our own supply, obviously an enormous and impressive armory.
1:08:12
Uh what is the what do you think the the PMC of the future looks like?
1:08:18
Right, you have warfare is changing.
1:08:21
We have autonomous systems now.
1:08:23
A lot of the new capabilities being brought online are, you know, drone based.
1:08:29
I imagine your sort of vision of of how you counter that uh probably ends up looking like a mirror to some degree, but but I'm curious.
1:08:39
Uh I'm sure you've thought about it more than anyone else. Uh yeah.
1:08:44
Well, you look back in history, the contractors on the battlefield have always been there.
1:08:50
And if you were laying siege to a city in Europe a thousand years ago, the people you contracted for, the PMC's would have been the guys building the trebuchets or the battering rams, which was at that point the most sophisticated system, mechanical system on the battlefield.
1:09:04
Um, now those that cutting edge, bleeding edge technology is obviously in the drone and autonomous space. And yes, you're right.
1:09:13
the the private sector can adapt and operate so much faster including in the cyber space as well that the that is how you deliver significant punch above your weight by having a few people that do those systems very very well.
1:09:32
That's something that we're still the US DoD procurement swamp is still truly not learning from fast enough is you go from idea to prototype to field testing in combat on in Ukraine in a matter of days or weeks and you get a buy decision from a customer that needs it.
1:09:53
That same process in the US takes years.
1:09:55
And so the obsolescence of ideas is so quick that we are missing uh and so that that speed of advancement is obsolescing I would say hundreds of billions of dollars of past procurement decisions now and and we're diluting ourselves if we think we're we're fat, dumb, and happy just because we've spent trillions of dollars over the last 25 years in stuff.
1:10:21
Buying stuff does not equate to buying capacity. Yeah, that makes sense.
1:10:26
Have you been happy to see the the Army's recent modernization efforts?
1:10:31
They actually came on our show and and talked about it for 30 minutes a couple weeks ago.
1:10:35
Seems like they are extremely motivated. Yeah.
1:10:40
And I imagine on your side these are maybe things that you've been pushing for for coming up on a couple decades now.
1:10:46
But what was your read on it?
1:10:48
Yeah, I I sorry I didn't see their specific comments, but I I know there's certainly trying to be a refocus on le on merit and lethality and not all the other distractions.
1:10:58
I would go as stream as extreme as it would sound.
1:11:01
I would push the purchase decisions of a lot of the next generation equipment down to the division or even brigade commanders instead of effectively what we have now is a socialist monolith buying apparatus of these PEOs through a few verticals back in Washington with enormous headquarter staffs.
1:11:22
And I I lauded the guys that are trying to reform the Pentagon.
1:11:25
But the fact is you have 800,000 Pentagon civilians.
1:11:29
If they cut that number to 200,000, now we're getting somewhere because you have so much dead wood that it literally blocks any kind of the speed or innovation.
1:11:39
And when I say devolve purchasing decisions, yes, that will come with with supply chain chaos. You're right. I want that.
1:11:47
I want the 25th Infantry Division based in Hawaii, which is probably never going to go to war alongside the First Armored Division based in Texas.
1:11:58
I want them I want I want them to buy and adapt current drone tech that's that to their unique needs of the theaters they're operating in and let the best ideas float to the top.
1:12:09
So then the Pentagon buys makes a much much bigger buy across it.
1:12:13
if they bought 10 or 15 systems that are okay but they don't really like the speed of advancement of obsolescence as we were talking about uh almost necessitates that kind of trial and error.
1:12:25
Have you been happy that uh the startup world is is you know firmly pro uh defense tech investing these days?
1:12:36
Uh is there enough new investment happening or is it is it already too much?
1:12:41
What's your read on on on just how much investment activity there's been in defense tech?
1:12:45
I am I am happy to see defense tech uh money flowing into those spaces.
1:12:51
I think there's a lot of great ideas.
1:12:52
My worry is that for the Silicon Valley crowd which is used to spending and investing at a pretty high burn rate um of overhead because they're used to finding somebody that will buy their stuff because they have thousands or millions of early adopters.
1:13:08
You don't have that in the Pentagon.
1:13:10
You have basically four buying verticals.
1:13:14
Army, Navy, Air Force, Marines, and maybe a little bit from SOCOM and a little bit from CIA, but that's it.
1:13:20
And so that that however optimistic or aggressive their um acceleration of purchasing timelines, it's not near what those defense tech investors are going to be used for used to.
1:13:37
And so the the amount of overhead burn time they have to plan for is probably three to four times what they're thinking it is.
1:13:45
And so I worry that there's a lot of great ideas that are going to wither on the vine.
1:13:49
Almost like planting a planting a seed in really shallow soil.
1:13:53
Until you get roots to where they can get to water, you're screwed. Yeah.
1:13:58
What's been your experience with uh the capital markets generally throughout your career?
1:14:02
Have you ever worked with venture capital?
1:14:06
Has private equity played a role?
1:14:06
Uh is there can debt be used to finance uh some of these problems that we're trying to solve?
1:14:14
Like h how have you interfaced with the capital markets?
1:14:16
And where has it kind of fallen short in your mind?
1:14:22
Well, I had a uh I had the luxury I because of my father's success, I was able to self- finance Blackwater. Sure.
1:14:30
And after the initial trunch of capital, I think it was six million to start the whole thing.
1:14:35
Buy the land, build the facility, uh we didn't have a huge capex load until the aircraft purchases started. Mhm.
1:14:44
But we went from one leased aircraft to 73 aircraft that we owned in six years. Wow. All self- financed. No other equity.
1:14:54
We got a a working line, a revolver. Mh. and and that worked.
1:14:58
Um, I really don't like partners. I don't like PE. I don't like boards.
1:15:06
I I don't like any of it.
1:15:10
I'm kind of a unilateral person that way.
1:15:12
And so, I'm not the I'm not the I'm not the right one to ask about that. Yeah, that's fine.
1:15:16
Uh, what was your pitch when you were scaling the team and hiring people?
1:15:20
Uh, I mean, I've heard the narrative that uh that the money's better when you transition into private military work, but uh was there more to it than that, or is that narrative kind of overblown?
1:15:32
Um, talk about building the team throughout your career?
1:15:40
that was it was it became a significant um vorticy a real center of gravity of being able to pull talent because um yeah you know seuite talent I I never gave away equity in the business because it was dangerous it was volatile and I paid I paid extremely well I paid very competitive bonuses and then would bonus two three to five times base at the end of the Mhm.
1:16:09
So it was effectively a big profit share and maybe that's not as tax efficient or whatever but um the people liked it.
1:16:16
The guys that you know our ideal staffing then was somebody that spent time in a special operations or a military unit.
1:16:24
Uh not a career, not usually a full career, no colonels, no generals ever um intentionally because they had spent so much time.
1:16:36
This is take note DoD or venture capital startups, people that spend 20 and 30 years in the military tend to get stuck in that way of thinking. Mh.
1:16:47
And so they can't think of just finding a a unconventional, different, cheaper, better, faster way to do something because that's just the way it has always been.
1:16:59
Um the um but our pitch to people was and especially for the guys deploying was you get to do what you're really good at and we will pay you very well for that every day you're in the hot zone.
1:17:11
The day you leave their pay goes to zero.
1:17:13
So we really our our pay for deployed guys is really more like paying a rough neck.
1:17:20
A guy that's going to a rig they're busting their ass for 60 to 90 days and they're off the rig. They're relaxing.
1:17:26
they're going to the beach in Thailand or or managing their family farm or whatever.
1:17:30
But they get to do they get to maximize their marshall skills whether it's fly a helicopter on night vision goggles at rooftop level at night or a a sniper, a instructor or whatever that might be.
1:17:45
they get to maximize that skill set and uh and get paid for it without the nonsense and and the layers of of stupidity that sometimes goes to the military.
1:17:53
In in in this like onboarding process as you're hiring, was there a story that you found yourself coming back to and telling uh people that would join the organization to highlight heroism or sacrifice?
1:18:06
like what what was the defining story that you like to tell to kind of characterize the work?
1:18:18
Well, I guess there's two.
1:18:18
One happened to me uh when I was speaking at the National War College in Washington.
1:18:23
So, there's 300 colonels that are training to become generals.
1:18:29
And they had me come in and talk about what we were doing as a PMC in the battle space.
1:18:33
And uh at the end of it, this colonel came up to me and he had just been in brigade command.
1:18:39
So in command of about 4,000 people in Baghdad and he said, 'I want you to know that at you know at the top of the dashboards of his people in Baghdad's Humvees were the Blackwater call signs and frequencies because that those guys knew that if they ever got in trouble, the Blackwater guys would come for them.
1:18:59
No bureaucracy, no nonsense.
1:18:59
And uh that was a I was happy to hear that.
1:19:02
That's exactly the guidance I gave my people was the good Samaritan rule applies.
1:19:07
If you see a friend in need, go help them out.
1:19:13
And uh and they did that above and beyond way too often.
1:19:15
And then I remember um at the change of command for Admiral Eric Olsen, who was the first SEAL fourstar ever and a friend and a great guy.
1:19:28
um a another SEAL admiral uh who was in charge of all special operations in Europe and Africa came up to me and he gave me a bear hug and he said, "Uh, thanks.
1:19:39
You guys really saved our bacon last week and you have um it was some really sporty flying."
1:19:44
I said, "That's why we have very sporty pilots."
1:19:48
And what had happened is a so there was a army special forces team active duty training up in Mali.
1:19:53
This is before the coup and the jihad, all the rest.
1:19:56
And their tents were staked down and a big wind, okay?
1:20:02
They called a shamal wind came through and it basically picked up the tent and tossed it with guys inside and it killed one, broke another guy's back, another guy's uh limb, trashed all their comm gear, and it gave another guy massive internal bleeding.
1:20:18
That was the guy that was in danger.
1:20:20
It actually picked our aircraft up off of its moorings, damaged the landing gear and the wing tip and uh our satcom was the only thing that was working in the aircraft.
1:20:30
So they called the other aircraft which was a thousand miles away and we're in the and we need medevac.
1:20:37
Uh and the shamal winds were still going and now it's getting to to nightfall and at that point the DoD would not let us fly in that area with night vision goggles even though we'd requested it. But that's okay.
1:20:46
Okay, our pilots came anyway and they landed on a single set of headlights in the desert at night.
1:20:53
Now, if you've ever flown over the desert, you see how utterly dark it is.
1:20:57
And so to land on a single set of headlights is extraordinary and extremely dangerous. Wow.
1:21:02
And they landed and our medic, our 18, our former 18 Delta medic was strapped into the back and he gets all these wounded guys on board and he starts to work on this guy who's now his blood pressure has gone to like 50 over nothing because he's gone through every bit of liquid and he works him for the next 36 hours keeping him alive.
1:21:25
Now the guy's body was distended because they couldn't figure out where the internal bleed was.
1:21:30
They get him back to Aagadoo and he goes way beyond protocol to save his life at at personal licensing risk to the guys, you know, to himself.
1:21:42
He opens him up on an operating table of this lousy Wagadugu Bkinaaso hospital and sure enough finds that the guy's spleen is ruptured.
1:21:52
That's where the leak was.
1:21:54
And so he removes it, clamps it off, puts it in a Ziploc bag, puts the guy's organs back in, sews him up, and then finally the high dollar Air Force surgical ward finally arrives 12 hours later. Too little, too late.
1:22:10
But our guy our guy completely saved his life and he took massive personal risk to do that because his job is to be a medic.
1:22:18
It's not to do surgery, but he did it anyway.
1:22:22
And so that's in keeping with the attitude of having the right people with the right attitude and the right skill set uh with with it.
1:22:30
And we try to do intentbased leadership.
1:22:32
Go forward and do good things and support the customer.
1:22:37
Make sure the good guys win. And that was a good day. Incredible story. Great story.
1:22:41
I want to switch gears completely and ask you about a topic uh that I'm sure you spent some time thinking about, but AI safety uh that you might you might laugh a little bit about that, but I'm curious if you've worked with any uh if any groups kind of globally around this idea uh you know just sort of the idea of AI safety as AI systems advance.
1:23:06
Uh, you know, the labs talk about this quite a lot, the potential dangers of misaligned AI and the dangers of a adversarial government with with uh advanced AI.
1:23:21
What uh what if anything have you uh are your thoughts on on that sort of very broad question?
1:23:31
Well, I guess um the first question comes back to who's Yes, I agree.
1:23:34
comes back to who's Yes, I agree. Those are all very concerning topics and undoubtedly there are hostile unfriendly governments that have done serious germ and biowarfare uh research and development that we're not aware about not not aware of but I
1:23:51
even question our own government uh when I think it's legitimate questions to ask of NIH and the rest of those guys for the gain of function testing that they were doing and funding in China which ended up causing apparently uh the COVID debacle. Um I'd say almost
1:24:07
Um I'd say almost like the hypocratic oath, let's first do no harm and let's not spend money on dumb people that should not be doing that kind of research in the first place.
1:24:19
Uh but you're right, it I'm I am certainly no AI expert.
1:24:22
I'm probably more of a u I don't know what I'm an expert in, but I I a few things.
1:24:32
I it's safe to say that if if AI can be used for malign purposes and it will be it's going to be done or it will be done.
1:24:40
So whether it's for um breaking passwords or uh or for targeted people I'm sure some some entities are well along in that development and uh and one should be careful. Yeah, that makes sense.
1:24:54
Um, I also wanted to ask you about there was uh some unfortunate news or at least it seemed unfortunate that Taiwan had decommissioned their last remaining nuclear reactor and are now completely dependent on imported uh natural gas.
1:25:13
I'm sure you have some strong opinions around this.
1:25:15
Yeah, I find that extraordinarily stupid of them. And that's a DPP.
1:25:20
I mean, look, the two political parties are the King tongue and the DPP. Mhm.
1:25:23
The KMT is the one that's really much more pro- mainland CCP tilted.
1:25:28
The DPP is just kind of a super green and a bit woke as a party.
1:25:35
Uh that they would carry through on this and turn off a nuclear power that was a nuclear plant that was not dependent on hydrocarbons coming from the Middle East is extraordinarily shortsighted.
1:25:45
uh it speaks to their uh disbelief, their lack of fully understanding that that we are not there to hold their hand and to guarantee their safety.
1:25:57
Um I have advocated as loudly as I can that the only answer for deterrence there is an actual hope guard because that government is corrupted at many levels.
1:26:13
any weapon system of significant or strategic value is already been probably pre-registered to be hit with not one but probably five Chinese missiles.
1:26:23
So whether it's an airplane, a submarine or whatever, the only thing that can inject uncertainty into an invader is a home guard.
1:26:32
Um in the same way that uh you know the American Revolution, there was only 3% of our population that took up arms to fight.
1:26:40
If we found and in Taiwan that's 24ish million people so over 700,000.
1:26:44
If we were allowed to recruit and train a home guard three to four weeks their military is largely useless.
1:26:53
Their reserve forces not so much. They don't.
1:26:57
Their idea of combat training is far from what ours is.
1:26:59
Um, but if you find the citizens that are willing, it injects a massive uncertainty into an invader's mind because the terrain itself, either super urban or mountainous jungle, uh, definitely favors the defender.
1:27:14
And if Xi Jinping and the CCP are going to go for it, they have to be quick.
1:27:19
they they themselves cannot afford a long multi-year fight like you see in Ukraine because they depend on hydrocarbons and um even food for imports.
1:27:34
Uh switching gears a little bit again, I'm curious uh to get your thoughts on the European defense tech ecosystem broadly.
1:27:42
I'm sure you've done a bunch of business out there over the years.
1:27:46
there's been some renewed, you know, new kind of spending bills.
1:27:49
Ryan Matl has been ripping fantastic stock, but I don't know about the company and and and I guess there's just this overarching idea that uh America has and now every European country wants to kind of own their own defense stack and there's a lot of energy there.
1:28:07
But how do you think that the the landscape around defense in Europe is changing either on the the government side or the private sector?
1:28:18
Uh the only country I see being serious about it at all is Poland.
1:28:20
They have invested they have invested and built some real combat capability.
1:28:24
Uh the Baltics have not yet. Finland a bit.
1:28:31
There's lots of talk about a big European defense uh extra spend.
1:28:33
Germany has been atrocious. I don't know.
1:28:38
I I think I think people have bid up Rhymatal because I I don't know that their earnings have been that spectacular.
1:28:45
Their stock price has probably done great. Yeah.
1:28:46
But that's I hope that's a hope of future earnings. Sure.
1:28:48
But um uh I just see a complete lack of seriousness of European political leaders uh to address their own defense and uh it's taking him a long time to wake up from this uh this hangover of of and lethargy.
1:29:09
That being said, that being said, the the most capable military in in Europe right now is the Ukrainian military.
1:29:14
The lessons learned that they have are very significant.
1:29:20
The drone tech is far and away the best.
1:29:23
Their ability to to fight against and to even conduct electronic warfare and even close air support in this environment is uh is leaps and bounds ahead of even what the US military is.
1:29:37
So that's the military to learn from.
1:29:40
Uh Ukraine does does have a corruption problem.
1:29:43
I hope sincerely that Trump is able to to get a ceasefire in place and to stop this killing because it's it's absolutely pointless.
1:29:52
It's just Slavs killing Slavs at this point and it's it's nobody's going to advance.
1:29:57
Um and you have a um a blend of old and new.
1:30:01
I mean, if you look at the pictures of the front there now, it's almost indistinguishable from the Battle of the S, right?
1:30:10
artillery duels, static lines, bunkers, all the rest.
1:30:14
Now, the problem is somebody can fly an FPV into your bunker on the other side, but between tens of millions of landmines, which make armored breakthrough very difficult, it slows down any attack so that the FPVs and and artillery can get to it.
1:30:28
You're not going to see any kind of blitzkrieg, Hanserian maneuver warfare there um until some significantly different weapon systems come along.
1:30:39
So look, Europe needs to get serious about it.
1:30:45
They're far from it at this point. Yeah.
1:30:48
I I at home these conflicts have been raging for years, but they feel very abstract because in America you just kind of go about your day and go to the grocery store and things feel normal even though uh the the global tenor of great power competition and and uh even these more minor conflicts is growing every single day.
1:31:09
um between Ukraine, Russia, Israel, Palestine, India, Pakistan, uh Taiwan, China, what is the most where should the focus be right now?
1:31:19
How should people stay up to speed?
1:31:21
Um how do we get an even an idea of of where these conflicts are going?
1:31:28
How should people um understand the nature and the risk associated?
1:31:35
Well, they can follow my podcast because I talk about these conflicts almost once a week. Okay, that's great.
1:31:38
That's a great place to start.
1:31:40
What's What's the name of the podcast? Let's plug it.
1:31:42
It's uh it's Off leash with Eric Prince. Fantastic. Yeah.
1:31:46
I got one last uh I got one last kind of question for you.
1:31:51
Uh the rumors that we've heard around how the whole Tik Tok debacle is being solved haven't given us a lot of hasn't gotten either of us excited.
1:32:05
Uh, are are we the you think we're the laughingstock of China that we can't, you know, ban the the the Chinese AI brain rot app from our app store?
1:32:14
And and will you promise us that Tik Tok will not run on your phone?
1:32:23
Uh, I don't think it's capable to run an unplugged phone.
1:32:25
Uh, but now that you mention it, I will make sure that is the case. That's great.
1:32:28
Um, but yeah, any other color on the Tik Tok ban or just general relationships between China and the US with regard to technology and business, which is our focus here?
1:32:41
Look, the Trump administration is still in the process of getting its people in the seats and getting those institutions organized.
1:32:48
Um, I think I mean what what the president just did today in showing video to Sir Ramaposa from from South Africa about the the ongoing murder and genocide of white farmers was absolutely fantastic.
1:33:02
And it I even happier that Senator Secretary Rubio is right there supporting the president. Sorry about that.
1:33:15
uh that it's it is a different administration than we've seen in the past and I am happy to see those kind of issues being addressed.
1:33:22
I wouldn't discount the tis tik tok thing being addressed properly this time since it wasn't the last you know the 45 administration. Yeah. Yeah, that makes sense.
1:33:32
Well, we're hoping for something good.
1:33:33
Uh but thank you so much for joining us.
1:33:35
This is a really fascinating conversation and fantastic stories really.
1:33:38
Uh I mean just I was sitting back just like I'm listening to the podcast. I'm listening to you. It's great.
1:33:45
Uh, but thank you so much for stopping by.
1:33:46
We'd love to have you back and we'll talk to you soon. Yeah. Cheers, Eric. Thanks for coming on. Cheers.
1:33:51
Have a great rest of your day. Bye.
1:33:52
Uh, next up we have surprise guest.
1:33:55
Not on the announcement post that we put out on X, but we got Anthony Papliano. Good friend of mine.
1:34:00
Hung out with him in Miami a year or two ago. We talked YouTube.
1:34:03
We talked all sorts of video essay stuff. And he's got a spa. He's got a spa now.
1:34:08
and he's also launching a an asset manager with like a trillion dollars under management already.
1:34:13
So, uh he's off to the races left and right.
1:34:17
Big week, big week for Pomp.
1:34:17
Uh fantastic show as well.
1:34:20
So, uh excited to invite him into the studio and catch up with Anthony Pompiano and get a bunch of great takes. You're here. Welcome, guys.
1:34:30
I love to see the uh the suit, but we need some ties going on. I know.
1:34:35
Caught us caught us lacking. Caught us lacking. We're working on a tie.
1:34:37
We're working with a on a tie that has ads on it.
1:34:40
That's that's our next ad unit. We'll send you one.
1:34:42
Thank you so much for joining. Uh how are things going?
1:34:45
Uh what's latest in your world?
1:34:47
How are you spending your time?
1:34:49
Because you got a lot of a lot of different things going on.
1:34:51
Yeah, look, I think that the future is very similar to what you guys are doing, right?
1:34:54
You got to be able to operate businesses.
1:34:55
You got to be able to invest capital and you got to be able to uh create content on the internet.
1:34:59
And if you think of what a business really is is you got to build products, you got to be able to uh invest resources and then you got to be able to market those things.
1:35:05
And so I don't think there's any different than what people used to do in the past.
1:35:08
It's just now that you do it more as kind of an individual brand rather than a company.
1:35:12
And that's what we've been doing. Yeah.
1:35:14
What what was the the the kind of the I feel like you're going through like the V2 of the Pomp media strategy.
1:35:18
The there's a daily show now.
1:35:22
Uh and you were building businesses before.
1:35:24
What did you learn from those and where is this going with the next uh iteration?
1:35:30
Yeah, I mean look actually we we used to have a daily show uh very similar to what you guys are doing.
1:35:34
what you guys are doing. uh the best business show which was hilarious when people would have to cite us breaking news you'd have to write the best business show crediting uh good nice little troll there but no I think you guys are doing it the right way right it's like everyone is upleveling and the way I describe it is when content started on the internet um everyone was
1:35:51
basically on like their webcam and then they were like in their kitchen in their webcam and then it was like oh I have like an extra camera instead of the webcam now we pretty much just think you need TV quality with digital distributionally and so everyone is leveling up you guys have built Yeah, I think you guys just got a 4,000 square foot facility, you know, all this stuff, right? We've got obviously studio stuff
1:36:08
We've got obviously studio stuff and uh it's just the the cost of the technology is compressing and now with a very small team, you're able to have massive reach and I think that's just going to be the future of uh all this content.
1:36:19
Can we just glaze X for a little bit?
1:36:22
I know you've built a business on X. We love X.
1:36:23
We talked to the team earlier this week.
1:36:25
I just think that like it's so underrated as a distribution platform.
1:36:29
people think YouTube and there there other great platforms, but the the the current moment with X has been particularly good for us.
1:36:37
And I wanted to know what your uh what your take on X is, how it's been going, and any uh any learnings that you have to share.
1:36:44
Is Linda hiding behind your camera or something?
1:36:47
Got a gun to Yeah, she's got a gun to her head. Yeah.
1:36:51
No, we're gonna we we we just invited Linda on the show. Amazing. Amazing.
1:36:53
Um, no, look, I I think that every platform is very unique in what it provides, both in terms of the type of audience.
1:37:01
Obviously, you know, who's on LinkedIn is very different than who's on X.
1:37:03
Uh, LinkedIn finds out things maybe a week or two after people on X find out about it.
1:37:08
Um, but also, I think the type of content that lives there is really important.
1:37:11
Uh, both in the form factor and also the actual substance of it.
1:37:13
And so, it it doesn't make a lot of sense to go and talk about, you know, kind of daily news on LinkedIn.
1:37:18
It makes a ton of sense to do that on X. Yeah.
1:37:21
And so what I think many people, including ourselves, have figured out over time is that we were essentially using X as an engine of growth to drive people elsewhere.
1:37:30
Go watch on YouTube, go listen to the podcast, go read the email, go to this website, you know, whatever.
1:37:36
Why not be able to just bring that content right there in the feed?
1:37:40
And I think that that's really what they have leaned into and that's why it's working.
1:37:42
And so the daily show that we have with them, I literally sit at my desk, talk direct to camera, we cut it up in this modular format and deliver it right in your feed.
1:37:50
You don't have to go anywhere.
1:37:51
you just consume it right there in these bite-sized components.
1:37:54
And it's no secret that your guys kind of clip strategy and the live streaming, everything you're doing is like very excentric.
1:37:59
And I think that doesn't mean you give up on the other platforms.
1:38:01
It just means that you're building for X and everywhere else is kind of the exhaust of that uh content. Yeah. Yeah, it is crazy.
1:38:09
Like there was all this fallout from like, oh, they banned links.
1:38:13
I can't link to my YouTube videos.
1:38:14
And I'm like, but you can just upload those videos here.
1:38:16
upload those videos here. like like they literally have a video product like like and oh like you you you wanted to link to text like you could just put the text on X and it'll do way better and it's not that complicated and yeah you might have to change your monetization strategy a little bit maybe the rev share is going to be different on
1:38:33
YouTube it's like a per video thing and X it's monthly but like it all works out you're just playing for attention and reach and obviously X has this the thing that I like about the flywheel is like is like on YouTube I could every once in a while another creator like you would comment on my video or I'd comment on your video and we'd see each other. But
1:38:50
But we couldn't DM each other on YouTube.
1:38:53
But on X, it's like if somebody sees a clip and they respond, "Oh, I think you know Pump's wrong about this or I love this."
1:38:59
Like we can all hop in a DM group chat immediately and just start talking and then hop on the show, do whatever.
1:39:04
And so like I think the water cooler with all this stuff is actually really really uh inventive.
1:39:08
Anyway, Jordy, you got No, it's been it's been interesting.
1:39:11
people, you know, will point out they're like, I don't understand how TBPN is like has has so much attention.
1:39:16
The YouTube channel, the YouTube channel is small.
1:39:20
It's like, yeah, we don't really spend any time on YouTube.
1:39:22
We were happy to just focus on one platform as, you know, a very young, a ton of important uh tech people hang out. Yeah.
1:39:29
Uh anyway, I I let's talk about the other news.
1:39:33
Uh the spaxs in the news.
1:39:33
How do you raise $250 million?
1:39:36
Where's that money come from? That's so much money.
1:39:38
I I mean we Jordy and I both raised money at various times, but like it wasn't six months ago that I think we were on the show talking about how there's nothing intrinsically wrong with back. Totally.
1:39:51
It's just been maybe misused or abused in the past and also just bad timing, right?
1:39:56
Because timing timing and and ultimately what matters is the underlying asset, right?
1:40:00
Are you you know so give us the pitch for your spack and then we're going to pitch you spacking TBPN. All right.
1:40:06
spacking TBPN. All right. Uh I mean look I I do think it six months is actually an interesting time frame because uh that's about when I decided hey I I should go do this and you know look I got the big audience for years bankers
1:40:18
have been coming to me saying do a spack do an RTO do all these things in the public market and I always joke that we're all about the same age pretty much everyone in our generation watched the social network and wanted to be Zuck or invest in Zuck. So everyone ran to the
1:40:28
So everyone ran to the private markets it was like you know Wall Street and hedge funds all that stuff is kind of loser stuff now the cool stuff is in tech.
1:40:34
Um but we have a world where this undisiplined monetary and fiscal policy is a huge persistent tailwind for liquid assets.
1:40:40
Crypto and Bitcoin obviously saw benefit from this gold public stocks etc.
1:40:45
And so I think a lot of people you're starting to see them say, "Hey, how do I marry technology stuff with liquid assets, right?"
1:40:51
And so you're you're going to see that um with the Spack in particular, I agree with you guys that the vehicle itself is not the problem.
1:40:57
It's that people were basically doing public venture capital, buy a money losing company with a high valuation and you got to try to grow into it.
1:41:02
And some people did do that, but a lot didn't.
1:41:06
And so it got a bad name.
1:41:06
But what I really have talked about is like look if you the other thing sorry to interrupt you there's an adverse there was an adverse selection issue at the time as well which was that the companies had huge access to capital in the private market so if they were spacking it was possible that they had exhausted yep you couldn't do the stripe thing and just billion tender you would spack or something.
1:41:28
Yeah but but there are businesses today that are fundamentally good businesses that have bright futures that maybe should just get public and get remember years ago talking to a public markets investor and he was like, "Well, yeah, like once you're in the public markets, like you shouldn't ever need to raise capital again because you're supposed to just be a going concern."
1:41:46
And I was like, "Wait, what are you talking about?
1:41:48
Like businesses lose money. I'm in venture.
1:41:49
Like this is what you do."
1:41:51
And he was like, "No, no, no.
1:41:52
Like, like if you're publicly listed, you should not be losing money for a long time because then you just shouldn't be a public company."
1:41:58
I was like, "Oh, I didn't even think about that."
1:41:59
But yeah, anyway, sorry to cut you off. Yeah.
1:42:02
Well, I I think that um you know, I've always looked at this as if you buy a good company that has all the right criteria to be a public company, um and you do it at the right price, which is an important component.
1:42:12
Basically, if you're a disciplined buyer, then it doesn't matter who the sponsor is.
1:42:15
It doesn't even matter what the vehicle is, right?
1:42:17
It should do well in the public markets to your guys' point.
1:42:19
Um and people weren't doing that.
1:42:21
And so, I think that that is kind of one big component.
1:42:22
The second thing is a huge part of what we specifically have said uh publicly and in the filings is we want to buy a profitable company.
1:42:28
So, we don't want to buy a company and then try to make it profitable. Like, that's hard.
1:42:33
Let's go buy a company that is already profitable.
1:42:35
And the reason why that 6 month mark was so interesting that you mentioned that is that's about the time I started to get lots of inbound.
1:42:41
And I frankly don't know why.
1:42:41
I think some of it was the political change in DC.
1:42:44
I think a lot in financial services there was like a regulatory shift from a headwind to a tailwind.
1:42:48
So, people are like, "Okay, like this may be a better time to go into the public market."
1:42:52
And then also like think of all of our friends who have built companies that started sometime between like 2015 and 2020.
1:42:58
those companies are just now becoming big enough to actually be public companies. Yeah.
1:43:02
And so there's like that cohort of people are now like ready to get into that public market and so they're looking for the vehicles to be able to do that and we hope that you know we're one of many that they consider to uh to be able to do that.
1:43:12
How does the actual deal work once you find a company?
1:43:15
Are you buying a minority stake in the company or are you buying the whole company and then the the as I remember it was like you'd usually see like a couple hundred million dollars in a spa but it'd be trading at four billion and then you take a company public and then the share price goes from there. Is that roughly correct?
1:43:31
I I I I don't know enough about that.
1:43:33
The best way that I know to describe it is like the spa gives you a lot of optionality which is why I'm attracted to it.
1:43:38
So you we raised 250 million into the spa.
1:43:40
um you have the ability to buy a lot of a company or you can buy a very little bit of a company, right?
1:43:46
But you have $250 million of cash.
1:43:47
You do have to go to those investors and say, "Hey, this is the deal that I found.
1:43:50
This is the deal that I negotiated in terms of the economic terms." They vote.
1:43:53
So if you bring them a bad deal, they can also just say like, "Ah, screw it.
1:43:57
We're going to take our money back and like, you know, good luck."
1:44:00
So you have to do a good deal, but you also have the ability to use pipes and other kind of capital markets tools so that you can size up in terms of total capital invested in one of these businesses.
1:44:11
And then the beautiful thing about this is it's a negotiation with a target company.
1:44:14
And so it's two people in the market coming together and saying this is what we think it's worth, right?
1:44:20
This is what they're willing to take.
1:44:21
This is what we're willing to buy.
1:44:22
And I think that that tends to if people are rational and disciplined uh actually end up with something better in the public market.
1:44:29
And it's different than a venture fund where the money is not literally sitting in the venture fund.
1:44:33
They need to make a capital call with a spack like the money sits in T bills, right? Is that is that correct? Correct. It's in the vehicle.
1:44:41
Um some people have interest, some people don't.
1:44:42
You there's some nuance there, but basically you can go to someone and be like I have the money.
1:44:47
Like I don't raise the money.
1:44:47
I don't but LPS can't tell me like hey sorry you know I'm not going to send the money.
1:44:51
like I got the money and uh and he's good for his 250.
1:44:54
So, so I I mean I'm super fascinated by this whole industry.
1:44:57
Uh and so are people pitching you?
1:44:57
Do you expect this to be something where uh you get an introduction from a venture capitalist or an investment banker or just a founder?
1:45:06
Is this like a founder deal or are you going to be going outbound finding some business where they're not even considering selling because they're profitable, they're happy and you kind of come in, give them the pitch and say, "Hey, this might be the right time to look at something else.
1:45:18
Let's let's do this deal."
1:45:18
Yeah, I mean there's a bunch of rules again in terms of like what you can do.
1:45:23
So right now I can't talk to any target companies and uh that'll be true for another 24- 48 hours.
1:45:28
Um but once the deal is closed, I can actually uh talk.
1:45:29
I've got a ton of people who have been reaching out to me.
1:45:32
Uh you know, frankly, some of them are probably interesting, some of them aren't. I don't know.
1:45:34
I haven't, you know, talked to them or done the work yet.
1:45:36
Uh but also I think that you have an idea of like, hey, it would be interesting to go talk to these types of businesses, right?
1:45:41
And so you want to go and and figure that out.
1:45:43
The whole key what what categories are are you excited about?
1:45:48
about? what is what is kind of you know what are you just going to kind of ignore right I I don't imagine you're going to like defense tech obviously doesn't make sense or or I would imagine doesn't make sense maybe only in partnership with TVPN no uh we're very focused on financial services um we do have a like a legal
1:46:08
mandate we could do things outside but financial services is a focus and the the idea there is um I think most people you know it's no secret right I've got a reputation from all the Bitcoin and crypto stuff and so like are you going to go buy that and I just look at financial services as it's all becoming one thing. You know, Robin Hood is a
1:46:20
You know, Robin Hood is a traditional exchange business, but they have crypto as a fast growing segment.
1:46:26
Kraken is a private crypto exchange, but they're adding US equities.
1:46:29
And so, we're just seeing the merging of the new and old worlds in finance.
1:46:32
And I think that's where you're going to continue to see this go.
1:46:35
similar to, you know, what we're building with Sylvia, that's what we're going to uh do on the spack side is we're just looking for um where can you get distribution directly to my audience with financial related type products, services, and companies.
1:46:48
Um and then that's where we want to play. Yeah. Talk to us about Sylvia.
1:46:52
I know you got 18 million sitting in cash.
1:46:53
We got some recommendations for hyper cars you could pick up if you need one, but uh break it down. Stable assets. Yeah.
1:47:01
Who who who before I tell P1's Well, before I tell you about Silia, who's the best sponsor?
1:47:05
Like, maybe we should just talk like, you know, you guys should just uh just boast them up a little.
1:47:09
Well, next up we got Wander.
1:47:09
So, we could talk about finding your happy place.
1:47:12
We sing Find Your Happy Place. Find your happy place.
1:47:16
Book a wander with inspiring views, hotel grade amenities, dreamy beds, top tier cleaning, 24/7 concier service.
1:47:21
It's a vacation home but better pump. Do it.
1:47:26
Just book a wander today. Do it right now.
1:47:28
We're not letting you let you could live in a wander forever.
1:47:30
People dream of organic integrations like that right there. We're gonna clip this. We'll both promote it.
1:47:37
It's a dream come true when Thank you for Thank you for participating in that with us.
1:47:40
Yeah, you're you guys got it.
1:47:42
Um, so Sylvia is uh Look, I I think every single person I know uh who is in like the business tech finance world, they got the spreadsheet and the spreadsheet is basically the thing that ties together all your accounts in one place.
1:47:54
So you know just like what do I own?
1:47:55
Yeah, there's not usually a lot of expense related stuff.
1:47:59
It's all about, you know, uh how much cash do I have, what are my investments, what are they worth, all that kind of stuff.
1:48:03
And so what we did is we built essentially a box that you go and you attach your brokerage accounts, your cash accounts, your crypto accounts, uh your private investments, your collectibles, people have added precious metals, guns, cars, houses, all kinds of stuff.
1:48:17
Um in that box, it gets encrypted and anonymized.
1:48:20
And then what we did is we overlaid a bunch of AI models on top of it so that you can do simple things like you can prompt it and say you know what's my debt to equity ratio or analyze my stock portfolio and tell me uh where you see risk but also you can email Sylvia, you can call her on the phone.
1:48:36
Um for those that are wondering, I do not have $18 million in my checking account and 500k in my savings account at this very moment.
1:48:42
That was some great that was some great dummy data.
1:48:44
But yeah, it's interesting because I feel like mint.
1:48:47
com was popular and kind of fell off and maybe even got shut down.
1:48:49
I I was getting an email about that. Maybe it's done.
1:48:52
Uh and so obviously this is a this is an opportunity, but it's very but those are very expense related, right?
1:48:59
It's all about like saving money and there is a huge market.
1:49:00
Obviously people in America, there's a lot of people focused on saving money.
1:49:04
But I do think that there is this very specific cohort.
1:49:08
You know, the the data that we have right now, we just crossed over $2.
1:49:11
5 billion dollars connected on the platform.
1:49:13
Um, it's been about four weeks.
1:49:17
You think about it, it's less than a thousand users that have connected accounts. Wow.
1:49:20
And so these are people who are multi-millionaires who are going onto this platform on average and connecting these accounts.
1:49:25
And so to them, it's less about, you know, how do I save money?
1:49:28
And it's more about how do I make more money?
1:49:30
And that's what they're really trying to figure out.
1:49:33
It's like where's my money? Where are my assets?
1:49:34
And then what can I do to actually grow those assets?
1:49:36
And we think that there's something interesting to build there.
1:49:38
We're going to let you go, but uh finish it out with where can people get started with Sylvia?
1:49:42
Uh well, after they go and they stay at a wander and they sleep on an eight sleep, then they can go to uh cfossilia.
1:49:50
com and they can check it out there. Okay.
1:49:52
Thank you so much for stop.
1:49:54
We'll have to have you back for a deeper dive. This is fantastic.
1:49:56
Excited to follow this back. We'll talk to you soon. Bye.
1:50:01
Uh next up, we're going over to OpenAI land.
1:50:03
Uh we got Nikkun coming in from OpenAI talking about a new release, new tools and features in the responses API, their MCP server.
1:50:11
We're gonna explain uh we're going to have him explain uh the the the adoption of MCP that standard and welcome to the stream. How are you doing? Boom. Well, yeah. Great to be here. This is exciting. Thanks.
1:50:24
Welcome to the temple of technology.
1:50:26
Welcome to the temple of technology.
1:50:28
Yeah, it happens to be this podcast specifically, not open AI.
1:50:32
not open AI. Also big big probably uh I mean big day in OpenAI Slack today probably a lot merger the acquisition that's very exciting acquisition um I'm sure we'll get into design and decision making around design and devices at some point
1:50:49
but uh let's start with the announcement today can you introduce yourself exactly what you do at OpenAI maybe even how long you've been there and then what you released today yeah sure my name is Nick I'm on the product team here working on our API been here for almost two years now. I've been working on the API since
1:51:04
I've been working on the API since day one.
1:51:06
Uh and uh mostly focused on getting out new models in the API, releasing new tools, just making the API much easier for developers to use to build applications. So yeah, that's me.
1:51:18
And then today, what what's the big announcement around tools and features in the responses API? Yeah, sure.
1:51:21
So u about two months ago, we released the responses API.
1:51:26
responses API. uh we we used to have uh an API called chat completions that was uh you know the the thing that almost all of uh AI was built on top of uh we we created a new iteration of that uh that endpoint called the responses API which is much more powerful it supports
1:51:45
built-in tools uh it supports multiple model turns it's really uh you know natively built for models like 03 and 04 mini uh which which have reasoning uh built uh into them And today what we did is we expanded the suite of tools that uh developers have access to. Uh so the
1:52:01
Uh so the most exciting thing is support for remote MCP servers.
1:52:06
And so now developers can plug plug in their responses API integration into any remote MCP server across the internet.
1:52:15
Um you you have you have servers like Stripe, Shopify, Asana, Linear, um HubSpot.
1:52:22
Uh, and I think this ecosystem is just going to, you know, grow so much more over time.
1:52:26
Really excited to see what people build on top of it.
1:52:28
Is the benefit of of direct MCP support something about latency?
1:52:33
The idea that because I imagine that I could be interacting with the chat GPT API or the GPT API and then and then make a separate call to Stripe's API and I could even have cursor windsurf write that integration for me.
1:52:46
And that doesn't seem that difficult.
1:52:48
And I've so I've always wrestled with with MCP being it's great, it's cool, it's a standard, but what's the real value besides saving a couple developer hours writing an API binding library or wrapper?
1:53:01
Yeah, I think there are a couple of angles to this.
1:53:04
So firstly, because these organizations and companies are building MCP servers, they're taking an LLM first approach to to to what the API needs to expose uh to the agent.
1:53:15
So, Stripe has, you know, a bunch of APIs that can be used to create a subscription or a customer um or a product or a price.
1:53:23
But for an LLM, they could just sort of combine that into a single function.
1:53:28
And instead of returning this like massive JSON object, they can return something that's very specific to the task that was being solved so that the LLM can like more easily understand what's happening.
1:53:40
So it's an opportunity to like rewrite your APIs to be very LLM first.
1:53:46
Uh so that's the first thing.
1:53:46
The second thing is like friction, right?
1:53:50
Like what you said is like maybe 2 hours of work, but why do the two hours of work when you can do it in like four lines of code in like less than a minute.
1:53:59
Uh and so this is just going to make it so much easier for developers to to put together multiple tools and whatnot.
1:54:06
Um, the other sort of thing that I'm really excited about is like responses API takes multiple turns.
1:54:10
So in chat completions, you would have to like put it in a for loop and like make it keep going until it tells you it's done.
1:54:19
The responses API, you give it three or four tools, you give it a task, it'll keep going automatically.
1:54:24
Single API call, finish the whole thing and give you the result.
1:54:28
So it's truly magical when you when you get that. That's amazing.
1:54:32
uh run famously wrote text is the universal interface.
1:54:34
Is it just a standing o ovation for him over there today?
1:54:39
Because this feels like a uh like a victory for the text as the universal interface paradigm.
1:54:44
But is there more nuance that you could add to that philosophy of of machines communicating with other machines via text?
1:54:54
Because that feels like, you know, again, the LLM understands JSON.
1:54:56
I understand what you're saying and probably faster and and and smoother, but at the same time, we could just train an LLM to speak JSON. Correct. Yeah, totally.
1:55:05
I think LLMs can understand JSON.
1:55:08
I guess the meta point there was uh if you could optimize your tools and combine them for tasks that you want AI agents and developers to be able to do, that's much better, right?
1:55:20
Like at the end of the day, uh a model is not deterministic.
1:55:21
So even if you ask it to call like four tools one by one, it it's going to have some instances where it takes a wrong step and then it needs to backtrack and like fix it.
1:55:32
But if you have these like curated functions and tools exposed via an MCP server that can directly go to, um it just makes it so much more reliable for agents to be able to like do the thing that they're designed to do.
1:55:44
Uh and so uh there's definitely some benefit to that.
1:55:47
Um in terms of text, yeah, that that that's uh that makes a lot of sense to me.
1:55:52
Uh uh yeah um it's uh that makes sense to me.
1:55:56
How how has your general high level view on M MCP evolved over the past you know few months?
1:56:04
Uh were you skeptical at first or were you always excited about the potential of it?
1:56:10
It seems like a total MCP victory uh at this point. Yeah. Yeah.
1:56:16
Initially, I mean, the the main goal was how do we get an ecosystem of tools started?
1:56:21
Uh, and OpenAI did this thing with plugins back in the day where we built it all on top of Open API specs.
1:56:29
Uh, and our plugins was uh integration was like a little too early for its time because the models weren't that good at like using these like third party tools uh really well.
1:56:37
Um, and we didn't take a second stab at it, but uh the MCP folks did. Uh, and MC MCP is great.
1:56:45
it it supports so much more than just tools. It um and it took off.
1:56:46
Uh it was initially like local only.
1:56:50
So you know there was some hesitation about like how will people easily plug into this thing but now with the streamable HGTV protocol so that these things can be hosted remotely across the internet uh is just perfect.
1:57:04
Like I'm so excited about uh what's going to happen with MCP over the coming months.
1:57:09
Um the hundreds of remote MCP servers that are going to exist.
1:57:13
developers are just going to like plug into them.
1:57:15
It's it's just going to make things so much better.
1:57:17
So really really excited about the opportunities here.
1:57:20
I mean currently most of these uh most of these like when you hit an MCP server you're kind of expecting a few seconds of of latency just because of how long it takes to inference these models.
1:57:32
Uh I've been kind of beating the drum on hey some of these models are just good enough.
1:57:37
Let's bake them onto silicon and make them respond in like two milliseconds.
1:57:39
Uh and I don't know if that's happening or not.
1:57:41
You don't need to comment on that.
1:57:43
I've I like midjourney V5 I was like it's good enough like just make it instant.
1:57:47
Uh V3 I'm like I'm ready for it just to be like instantaneous.
1:57:51
Certainly with deep research and and chatb generally. Um I'm super satisfied.
1:57:57
I'm like I don't really have any uh any requests for improvements to the models.
1:58:01
I just want faster and cheaper and easier.
1:58:03
Um but do you think that if inferencing does go through some sort of massive speed up uh whether it's from custom silicon or or architectural redesign does geocaching or edge computing become important to the way MCP servers communicate with each other?
1:58:27
for example, like I don't necessarily if I'm going to the Portland AWS hosting of some MCP server, I don't actually want to be routed to Virginia uh by accident.
1:58:36
And so do you think that that will require like a new rethink of the spec or is that something that can just be naturally handled by load balancing?
1:58:43
You know, the spec is just like how do you communicate between uh the MCP client and the MCP server?
1:58:49
So it tells you like you know literally the JSON RPC like object that that needs to go between the two parties.
1:58:56
Uh optimizing the geocaching of it really depends on the service provider, right?
1:59:04
So if I'm going to call mcp. stripe. com or allirds.
1:59:08
com/apimcp which is by the way allirds has an MCP server. How crazy is that? Okay.
1:59:12
Uh Shopify has to route it to like the right place. We need an MCP server. Help us build one.
1:59:17
I don't know what we do but uh we want to We want to get in on the action. Yeah.
1:59:21
We feel a little bit left out.
1:59:23
Create a create a create a Shopify store and you'll have one instantly. It's Oh, that's good. Yeah. Okay. Yeah.
1:59:28
For the merch, you can integrate. Yeah. Yeah.
1:59:30
So, I mean, OpenAI already does a lot of geo routing based on on on where the inference request is coming from.
1:59:36
So, we try to like route it to something near and then when when the request leaves OpenAI's infrastructure and goes to say Stripes or Shopifies, uh they could do that as well.
1:59:48
uh based on uh you know either like an explicit URL or something more magical on their end where they can detect where it's coming from and and route it to the right place. Makes sense.
1:59:59
Um any other questions, trends in pricing?
2:00:02
I know that there's pricing and availability updates. Anything to share there?
2:00:07
No, pricing is like uh you know it's it's an infrastructure play over here. Uh MCP is free to use.
2:00:11
You just pay for the tokens.
2:00:14
Uh we have code interpreter which is like the Python tool.
2:00:17
By the way, the Python tool is the thing in 03 that can do geoger like things when you when you upload your images.
2:00:23
So you can just bring that into the API. Yeah.
2:00:25
So why is Python relevant there? Yeah.
2:00:28
So what's actually happening is when you when you put a an image in the 03 model, it writes Python code to like zoom in. Sure.
2:00:41
Like try to find tries to find clues. crops them. It inverts the images. Got it.
2:00:45
It then like does web searches.
2:00:47
So, it's like all the all the image manipulation is based on Python and that is enabled by the code interpreter tool which is also available in the API today. Um, that's interesting.
2:00:57
I I noticed that because I was I took a picture of uh of a of a podcast.
2:01:03
I wanted to see how tall the table was.
2:01:04
I uploaded this images from the Pat McAfee show and and and I noticed that when I unpacked the 03 internal reasoning, it was like writing tons and tons of lines of image processing code in Python all to to draw lines and count pixels to see how tall the the the table was relative to the person.
2:01:22
It wrote like 500 lines, 5,000 lines of Python. We were joking about it. 403 yesterday.
2:01:25
I said, "How many white boxes in this picture?"
2:01:27
And it and it and it was coding for like almost 14 minutes.
2:01:32
It was like unfortunately failed.
2:01:34
It was like putting a putting a software engineer on the task for a month.
2:01:38
But you know now you get it in two minutes. Uh but it is incredible.
2:01:40
I can see where that goes.
2:01:42
It just adds a completely different angle of attack on reasoning and it needs to be an important tool in the tool chest. So yeah.
2:01:50
How are you how are you uh leveraging codeex today?
2:01:53
I'm sure you guys got got early access, but I'm I'm curious if and where it's fitting into your your toolbox toolkit today.
2:02:00
I see the engineers using it nonstop like uh we had to make a couple of changes for the launch.
2:02:08
We had to like fix error messages.
2:02:10
These are like such small things and and you can just fire up a codex task and do it.
2:02:14
I've used it for docs changes uh a bunch of times where I'm just like there's a typo here, change this heading and and these are tiny things.
2:02:23
I just like write it, forget about it, create a PR.
2:02:24
I mean that I don't have to bug people anymore.
2:02:28
It's it's really it's really helpful.
2:02:28
Uh and I think engineers are actually doing like even more complex tasks on top of it.
2:02:35
Um but I just see everyone in the company using it which is uh really cool. Very cool.
2:02:40
What are you uh what are you daily driving as a personal smartphone these days?
2:02:45
Is are you iPhone, Android, maybe something else? What do you got? I know. You got early access.
2:02:51
You got something in your pocket. Okay. Okay. No leaks.
2:02:52
No leaks today, but we've been asking.
2:02:56
We still part of Johnny's legacy.
2:02:59
We we we tried to trick you into leaking the hardware here, but we'll get you back next time. Yeah.
2:03:04
Well, I'm sure we'll find out soon.
2:03:06
But it's an exciting It's an exciting acquisition.
2:03:08
Uh, but thank you so much for stopping by. It's very exciting. We'll talk to you soon. Yeah.
2:03:11
Congrats to you and the team on the on the launch. Fantastic. We'll talk to you soon. See you. Bye.
2:03:15
Next up, we got David Sandra coming in the temple of technology, the godfather of podcasting himself. That's brother behavior. Brother behavior.
2:03:23
We need a lot of uh we need a lot of sound effects for this one.
2:03:27
Uh he's been on absolute tear dropping podcast after podcast.
2:03:30
But uh today we want to talk about John Johnny IV. Yeah.
2:03:36
And Apple and Steve Jobs.
2:03:36
What would Steve Jobs be doing in this situation?
2:03:40
I want to I want to play out that scenario for sure.
2:03:41
Um, and then we got Scott Woo coming in from Cognition.
2:03:47
Um, talking about Devon coding agents.
2:03:49
Every company has a coding agent now. Crowded market.
2:03:51
I want to know how it plays out.
2:03:53
So, we're excited to talk to him about that.
2:03:54
Then we got Joe Westhal from Bloomberg OddLotss coming on the show.
2:03:58
We'll get an update on the market.
2:04:00
It's been flat for a couple days. Rough week for Joe. It's been terrible. Yeah. Yeah. Yeah.
2:04:04
He's not getting his dopamine fixed.
2:04:05
How is he How is he getting through it? I want to know.
2:04:07
I'm sure he's finding obscure exchanges where he might be going 100x short volatility.
2:04:14
Yeah, maybe he's just printing.
2:04:16
Might be shorting Costco at 50 times. Maybe. We'll have to ask him.
2:04:19
Uh anyway, uh let's do some more ads in the meantime.
2:04:24
Let's tell you about Bezel.
2:04:25
Your bezel concierge is available now to source you any watch on the planet. Seriously, any watch. Any watch.
2:04:29
And I think we got it through our day. Download the app. Yeah.
2:04:34
Um, we can also go through some I'm actually shocked at how frequently I use bezel. Like I'm a DAU.
2:04:44
I was talking to Quaid Yeah.
2:04:44
friend and one of the co-founders and he was like, "Yeah, a lot of like a lot of people actually daily are just DAUs." Yeah.
2:04:54
Regardless if they're like actually shopping right away.
2:04:55
Well, we got David Senra from Founders Podcast coming into the studio. Welcome to the show. How you doing, David? What's up, guys?
2:05:05
Uh, before we begin, I have to apologize for not wearing a suit, but I think you'll make an obsession an exception because I'm wearing a ramp shirt. Oh, there we go. Let's hear it.
2:05:14
Switch your business to ramp. com. Save time and money. Um, how you doing? What's the latest there?
2:05:22
There used to be this thing I I mean, it's probably still a thing where David like would you would you'd be happy to meet somebody, but the one requirement was that they listen to founders.
2:05:32
Eventually, it's going to be you listen to founder and your business runs on ramp. I won't take it.
2:05:36
And if you don't check both boxes, not happening. No shot. No shot. That's hilarious.
2:05:42
Anyway, what's the latest?
2:05:44
Uh, did you did you read the Bloomberg article uh about Johnny IV joining Open AI? Okay.
2:05:50
No, you tell me what's going on.
2:05:53
I just got a bunch of text messages about this and I think there's something related to Steve Jobs coming back to Apple we could talk about too. That's interesting.
2:05:59
But no, I have no idea what's happening. Please.
2:06:00
what's happening. Please. So the article is obviously like it's amazing in the sense that it's Johnny IVive a $6 billion acquisition huge number 2% of open AI equity going over to Johnny Iive and his team but the and it's also super
2:06:16
interesting because they're going to build a hardware device for chatbt for for AI an AI first hardware and if there's any team that can break through with a consumer electronics product it's got to be Sam Alman and Johnny IV right like they are an incredible combination to go after this. At the same time, the
2:06:31
At the same time, the article basically said nothing.
2:06:33
And and they put out a 9-minute video and it's just them saying, "Oh, he's one of the most thoughtful individuals.
2:06:39
He's one of the best people to work with. His family's amazing."
2:06:41
And they're just talking about like how much they like love working together.
2:06:44
And it doesn't actually give us any advice on like or any guidance on like is it a phone, is it a watch, is it going to fold, is it big, is it small.
2:06:53
That's very Steve Jobsesque. He'd be very proud.
2:06:54
Uh, listen, I guess going into this like I I'm an unadashed like Johnny IV uh fan.
2:07:02
Okay, there's actually an excellent uh biography of him that I think everybody should read.
2:07:05
I read it like five years ago.
2:07:07
If you want to um listen to the podcast I did on, it's episode 178 before you you get the book, but I'd highly recommend reading the entire book. Um, yeah.
2:07:15
I just I think uh if you think about like the conversation we had this morning uh about taking things very seriously.
2:07:22
I facetimed you guys at like 6:30 in the morning and you're in your studio like getting ready and preparing.
2:07:29
You're sitting in these exact seats. Yep. Yeah.
2:07:30
Like I I think that's one of the things that jumps out in Johnny's uh biography is just like his intolerance for a lack of effort.
2:07:37
So, like there's a great story.
2:07:39
Um, I just pulled up a bunch of my highlights from um the book, but when Johnny was in design, he was studying industrial design.
2:07:47
Um, I think it was in England.
2:07:48
And there's this great line where a friend of his comes over, they have a project, right?
2:07:51
And they're making these like foam pro prototypes of a product they're designing.
2:07:56
And he goes in Johnny's apartment and he found over a hundred foam model prototypes where most students would build half a dozen.
2:08:03
So the normal expectation is, hey, I'm gonna make six prototypes of this, you know, new design for this product I'm doing.
2:08:11
And Johnny's like, yeah, that. I'm doing a hundred.
2:08:13
And I think when you read the book, you it makes sense.
2:08:17
Like, um, I just tweeted this out where it's actually Steve Jobs. Uh, let me pull it up.
2:08:22
Steve Jobs talking about Johnny.
2:08:24
And he's talking about Johnny with Walter Isacson, and he's writing the book with Walter Isacson as he knows he's dying.
2:08:30
And he calls, you know, he says that essentially if he had a spiritual partner at Apple, it would be Johnny.
2:08:36
He's not just a designer.
2:08:36
That's why he works directly for me.
2:08:39
He has more operational power than anyone else at Apple except me.
2:08:42
And there's no one who can tell him what to do besides Steve.
2:08:45
Um, so again, I always think of like actions express priorities, like one of my favorite maximums.
2:08:49
And Steve is telling you, you know, he had this probably the highest bar of talent out of anybody that's ever maybe anybody's ever existed. Yeah.
2:08:56
And he's saying, "This is the best guy. This is my number two."
2:09:00
Uh yeah, I think that like speaks volumes about like what he thought of in terms of uh Johnny's talent. Yeah.
2:09:04
I mean, I it's it's so interesting because uh I guess the o I guess the open question in my mind is does Johnny IV need a Tim Cook figure to do what he does?
2:09:20
And what I mean by that is that obviously Tim Cook's been the CEO of Apple for over a decade now, but he was a supply chain mastermind and Johnny IV could throw a crazy idea out there like an aluminum mil unibody design for the MacBook Pro and Tim Cook was able to actually go get it done.
2:09:42
And in the new era of global competition with China, faltering relationships and not a legacy at open AI around hardware manufacturing, which is a unique skill set, can Johnny Iive actually deliver a product that gets into the hands of millions of consumers with the quality that Apple is known for without a Tim Cook-like figure.
2:10:11
Will we see another acquisition?
2:10:14
Steve Jobs needed a Tim Cook figure. I think anybody does.
2:10:15
So like yeah, the answer is of course he does.
2:10:18
Like we'll have we have no obviously no idea what they're going to build.
2:10:21
Um what I think is just interesting is which way more applicable to like us and the people that are listening is just like to study I I just tweeted this back at Sam Par from my first million.
2:10:32
It's like you don't copy the what, you copy the how.
2:10:36
And so I think the important thing of like yeah, you could hypothesize of like what Johnny's going to build or you could just like, hey, how does this guy approach like his life and work?
2:10:42
Is there anything that he said in there that I could actually use?
2:10:45
And I think that's like much more interesting to me than be like, can you predict what's going to happen with this thing that they told no details about and that like who knows? I don't know.
2:10:55
Like can they build hardware? Who the knows?
2:10:57
Like we're gonna find out. That's the fun part.
2:11:00
Uh but yeah, I'm not into like prediction. I'm just like we'll see. Yeah.
2:11:04
So, I mean, he wasn't a founder at Apple.
2:11:06
Uh, he became a founder with by founding IO and, uh, his his love for his agency.
2:11:13
Uh, now he's back in the senior executive role.
2:11:16
Can you tell me a little bit about his management style, how hardcore he is, anything that ties to Johnny Ives like living like founder mode or any of those kind of uh, ties to other executives?
2:11:31
kind of what we might expect or what we might pattern match against.
2:11:34
Is it just Steve Jobs or are there other people?
2:11:37
I think he had like I would argue like you know equal to um sense of like a high threshold for quality that Jobs did.
2:11:44
I was just listening to the episode I did five years ago on Johnny's biography and it's like talks about it's like they they thought very similar about design and products and they were very much partners you know almost like a co-founder relationship is that I know that'd be controversial but like that's the way I would think about it like how deeply they were working together.
2:12:01
Um, but what I say is he's like much more diplomatic where Johnny's like, you know, we could do better than this and Steve Jobs like this is Get it out of my face.
2:12:12
And and part of that they I think they hypothesized his biographer hypothesizes that it's like, you know, his his growing up in England like his just more reserved and quiet manner.
2:12:20
Um, but what I loved is I think the very beginning of the book, uh, the the the author is interviewing him and he he was and this is even before he was going to write a book on him.
2:12:32
He was just covering Apple at the time.
2:12:33
I think he was a reporter if I remember correctly.
2:12:34
And he just like wanted a sound bite on this new Apple product and he says, "I just wanted a sound bite."
2:12:39
But he launched into a passionate 20-minute soliloquy about his latest work.
2:12:42
I could barely get in a word edge-wise.
2:12:44
He couldn't help himself. Design is his passion.
2:12:45
Um, and I think that's just he's just obviously completely obsessed and willing to like work hard.
2:12:51
There's another thing that um I thought it was interesting.
2:12:53
He calls it attention to invisible details.
2:12:55
Uh he says, "I've uh Johnny cared deeply about the parts users would never see, believing that quote care was taken even on details hard and soft that people may never discover."
2:13:07
And he says that he was fanatical uh he he he provided a fanatical care beyond the obvious stuff.
2:13:13
Um so these just like that's why I'm just a huge fan of anybody like this.
2:13:15
I don't care if you're building the iPhone or you're making a pizza.
2:13:18
Like just I I'm very attracted to and obsessed uh with people that take whatever they're doing very very seriously.
2:13:23
It's funny that the opposite is is Enzo who's like I don't care if the door panels align.
2:13:29
I care if the driver shits his pants.
2:13:35
That's I I would actually argue that they're very similar because Enzo remember like Enzo was building a race car.
2:13:42
He wasn't building like he wasn't trying to it just happened that when they were winning all the races then you have all these consumers say hey I want a Ferrari though too.
2:13:49
But I would actually think like what they both share in common is they keep the main thing the main thing. Yeah.
2:13:54
And in Enzo's cases he's not trying to buy build a luxury product.
2:13:57
He's trying to literally make the fastest car that he didn't give a about Ferrari the actual auto manufacturer if you read his biography.
2:14:05
And in fact, the longest book I've ever read for founders.
2:14:07
Let me see if I can find it real quick.
2:14:10
Uh was a thousandpage biography of Enzo Ferrari. Enzo. No way.
2:14:15
And it's it was just republished. Same author.
2:14:18
Uh this book is excellent.
2:14:21
It's Enzo Ferrari by Luca Damonte.
2:14:23
So this used to be a thousand pages.
2:14:25
No one bought the goddamn book because no one wants to read a thousand pages.
2:14:28
So they edited it down to 450 pages. Um yeah.
2:14:32
So, so if you you know Enzo's one of my favorite founders I've ever I've done like three or four uh Johnny IV at Lovefro his his design consulting agency.
2:14:41
One of their biggest clients is Ferrari today. I didn't know that.
2:14:45
And so there's probably some linkages here.
2:14:46
So So can you give us a little bit more about what we can expect?
2:14:50
We were joking about maybe OpenAI should just the device the device integrate with the Ferrari and and and and you just need to buy a Ferrari if you want to access the premium tier of chat GPT.
2:15:00
You can only access You should only experience it saddle leather of a Ferrari.
2:15:06
That that that that will only work if they're going to do a collaboration with TBBN.
2:15:10
I think that's a Oh, sorry.
2:15:12
I I I I can't do my homework today.
2:15:14
Uh my Ferrari's in the shop this this summer.
2:15:16
I'm going to I'm going to force you to go to the the Ferrari dealer in LA and uh get one because if you're really such an Enzo fan, you should have a Shouldn't you be honoring his legacy every time you go to the grocery store? It's funny.
2:15:29
I would I drove past the uh my daughter used to go the my the do my the school my daughter used to go to was actually next door to a Ferrari dealership in Miami which is like the most Miami thing ever, right? Super Miami.
2:15:40
But there's a great I tweeted it out when I passed by the other day because it says uh he had this great line where he's like uh passion can't be explained. It can only be lived.
2:15:49
And I was like oh that's like a really good uh way to think about the way he lived his life.
2:15:53
But when he finished that thought it's just like his primary thing was just like I want the fastest race car in the world.
2:15:57
And so that's what he meant.
2:15:59
and he's like, "I don't give a if the doors don't line up.
2:16:01
I just when the the race car driver hits the gas, I want him to his pants."
2:16:04
Which is also hilarious quote. Yeah.
2:16:09
I wonder what the AI equivalent of that is.
2:16:10
Like what is the Ferrari of AI models right now?
2:16:12
Because they're all kind of I'm personally I'm personally very excited about this.
2:16:16
Uh I was just talking to our mutual friend Patrick from Invest Like the Best and we both feel very similarly.
2:16:21
It's just like I don't want to look at screens all day long.
2:16:24
So I know you had um you had that guy that's building friend.
2:16:27
His name escapes me at the moment. Avi. Yeah.
2:16:32
Just this idea like I'm not saying I want to wear something around my neck or have glasses on my face, whatever, but anything that would kind of get me out off of looking at screens.
2:16:37
Like I'm I'm very uh like I would buy it day one just to test it out. Yeah.
2:16:43
John was saying like maybe a Walkman style device with wired headphones, you know, kind of analog but just piping the AI right into your brain. Maybe a football pack. It's interesting.
2:16:54
I I was thinking it's like Elon, what is Elon's AI hardware plays? One is Optimus, right?
2:17:03
But then the other is actually just Neuralink.
2:17:05
He's like, I want to jump, you know, he's he could say, I want to jump to like the end state, which is like human and computer become one. Pretty crazy. Yeah.
2:17:13
I I want the opposite of that.
2:17:16
Um I did this episode on Christopher Nolan and I it was very fascinating because this guy makes some of the most high-tech movies of all time.
2:17:23
They're very difficult to make and yet he lives a complete like analog life.
2:17:26
Uh one of the famous famous anecdotes from his biography was hilarious where he doesn't have an email, doesn't even carry he carries like a normal flip phone, won't use GPS, like goes to a new city and has to like feel his way around and ask for directions and everything.
2:17:37
But the funny part was he would people, you know, send scripts everywhere.
2:17:41
So they're like, "Oh yeah, just like I'll read your script. Email to me."
2:17:43
He's like, "I'm going to fly on a plane tomorrow and I'll be at your house at noon and he'll hand you the script if he wants you to be in the movie."
2:17:49
And then he sits there and stares at you while you read it and then you're done.
2:17:53
You talk about it and then he takes the script back and leaves. There's no other copy. Yeah, man.
2:17:59
Everybody's got to figure out their own integration for what they want for technology.
2:18:02
I want to disappear in the background.
2:18:04
Many cases like if you study this book and again I would order the book.
2:18:07
I would definitely read it. It's worth it.
2:18:09
Especially because it's going to play a huge role at one of the most important companies in the world right now. Totally.
2:18:13
Um but I I I just like I want a more like I want technology to disappear.
2:18:17
I do uh there is some quotes in the book that might be fascinating that kind of answers your question John.
2:18:22
I didn't mean to like sideskirt at all, but as far as like what he how he thinks about design, it says he takes big chances instead of evolutionary approach to design.
2:18:29
And if they focus grouped any of his designs, this is a description of him in college.
2:18:34
It was also accurate when he went to um Apple.
2:18:37
And if they had focused his designs, they would they wouldn't have been a success.
2:18:41
So he's not looking for like really wasn't interested in like any kind of like iterative design.
2:18:45
Uh he wants something that looks completely different.
2:18:49
different. And there is a weird uh story I think some people might find it odd story from his uh college days when he was studying design that he talked about that the best physical objects they they are very inviting to the touch that humans have this I forgot what he called
2:19:05
it like an obsession he has a word for it but essentially obsession with touching things and uh so he was asked to like design a pen and people like okay I'm going to write you know upside down or it never runs out of ink he's like no but what if people do a pence. Like
2:19:19
Like anybody holding a pen in their hand, they're like clicking it.
2:19:21
They they move it around.
2:19:23
He's like they want to I didn't think he said this word, but they want to finger it for lack of a better word.
2:19:27
So they like play with it a lot.
2:19:30
Um and so yeah, any kind of physical device that he makes, I think isn't going to just be like, oh, this kind of looks like an iPhone.
2:19:35
I think it's going to be like drastically different.
2:19:36
Yeah, it's a big challenge.
2:19:39
Um, and I mean he he's pulled away from Apple a lot of the people that were integral to so to the magic of of the devices, the MacBook, the iPhone, all these things.
2:19:49
I'm really excited to see it.
2:19:51
2026 can't come soon enough.
2:19:54
Uh, what about Jim Simons?
2:19:54
You did a recent show about him.
2:19:56
Uh, very different from Johnny IV, but what are the similarities?
2:20:02
Because you think about a hedge fund versus a consumer electronics manufacturer.
2:20:08
I imagine that there are no similarities and yet I imagine you found some.
2:20:11
How much would you pay to rip a heater with Jim Simons and Johnny once?
2:20:17
I mean that's a priceless heater right there.
2:20:20
That's potentially a seven figure heater.
2:20:25
This is why I think I got so many text messages and they linked me to whatever article today.
2:20:29
Uh because OpenAI essentially what bought Johnny's company.
2:20:32
Johnny was the founder and it's for like six billion or something I guess is the thing.
2:20:35
And everybody said the same thing in the text messages like how much money did Apple pay to like buy back Steve Jobs. Yeah.
2:20:42
And this is one of the thing that Jim has in common, Rockefeller has in common.
2:20:47
Bezos suggested Bezos shareholder letters which I think is a really good episode.
2:20:51
Michael Dell over and over again.
2:20:51
It's like this idea of like you you absolutely must work with the highest quality people you can and everybody says that and nobody does it because it is so hard to do it.
2:21:01
And one of the reasons I also uh Brad Jacobs talks about this like you have to overpay for talent because it's nearly impossible to overpay for talent.
2:21:08
So the obviously the canonical example of that is hey uh Apple's going to buy next for 500 half a billion dollars.
2:21:14
They paid essentially half a billion dollars to rehire Steve Jobs and they got the deal of a century.
2:21:19
Nobody's going to look back at what Steve did like, oh you overpaid it. It just doesn't happen.
2:21:22
So Jim, the crazy thing about Jim and then actually me and uh a mutual friend of ours that I can't name just stayed up till like 2 in the morning talking to this guy uh that works at Citadel and it was very fascinating like Ken Griffin was the same exact way.
2:21:38
We was like I heard some of the competition for his top people and guess what like they would blow I think it would shock how much money they make but then you look at their performance it's like direct essentially Ken won't cap their compensation.
2:21:50
So like if you make him more money, you're like unlimited money that you'll uh that you'll get because of he understands the the talent.
2:21:56
So the thing about Jim is that's that's nuts, right?
2:22:00
Is he churns through partners and you know people say Jim Simons had a hedge fund.
2:22:05
I don't think about as a hedge fund.
2:22:06
I think he had a magic money machine.
2:22:08
Like it was very different than normal.
2:22:10
You know, the the fund made well over a hundred billion dollars in profit and it continues to make from what I understand, you know, many many billions in cash flow every every single year and there's no outside money.
2:22:21
It's just owned the only money's in there was Jim's family's money and then you have to work at Renaissance. Yeah.
2:22:26
Uh and then if you quit they they you have such low turnover is because if you quit you can't invest in the medallion fund anymore. You're never gonna quit.
2:22:34
Go invest your money somewhere else.
2:22:36
But but the crazy thing is so what I'll do is like the fascinating thing is Jim is like churning through partners because it takes him decades of trial and error to to actually get the system that he wants.
2:22:48
But then he would talk about and he talks about this when he's an older man too.
2:22:52
Uh he'd give interviews much older because he wasn't doing that when he was younger.
2:22:55
But he's just like you have to work with the very best people and when you identify that person you have to just put all the effort you possibly can to convince them to work for you. Okay?
2:23:02
A lot of people understand that.
2:23:03
But then you go look up the people he identified and it's just like he hired nothing but mathematicians and physicists and you like yeah it's not just like some professor at in a math department somewhere.
2:23:14
It's like this guy has a theorem named after him and he had there's a law named after him.
2:23:18
It's like literally the top of the top of the top of the top um because they were you know made hundred billion dollars plus with 300 anywhere from what I've heard 250 to like 400 employees.
2:23:26
So a very very tiny team, small team relative to the profits produced of just unbelievable A players.
2:23:34
And I I have to imagine that's what you know people at Open A are doing.
2:23:38
It's like well who's the best in who's the best designer in the world?
2:23:40
If we're going to build our own product, who would you go and get and Yeah. Yeah. 2%. It's totally worth it.
2:23:48
That's what I was saying.
2:23:48
It's a big number, but it's a big company. It's a big opportunity.
2:23:51
And of course Josh uh Josh Kushner was already in IO.
2:23:59
So now he's getting so now he's getting uh more open Thrive gets more of Open AI.
2:24:06
But um but anyways, speaking of mathematicians, we got Scott Woo.
2:24:09
Yeah, we got Scott Woo from cognition coming in. I'm jealous. I love Scott. Big fan of his.
2:24:14
Have you guys ever played uh the video of him at a math competition when he's like 12 on the show?
2:24:19
I mean, it's been viral sometimes.
2:24:21
cuz we haven't played it on our show, but it still plays in my mind, but it's excellent.
2:24:27
We'll bring him in to the studio.
2:24:27
Thank you so much for stopping by. Yep, of course. See you guys soon. Love you. Thanks for having me. Talk to you soon. I love you, too. Bye. Love you, too. Bye.
2:24:37
Uh, and we'll bring Scott in in just a second.
2:24:39
Jordan, you want to kick off the intro?
2:24:40
Take a quick break and uh maybe do some ad reads, talk about his former employer, ramp. com.
2:24:44
Switch your business to ramp. com. John, use the restroom. Use the restroom. Get out of here. No dead air. No dead air. We do not stop talking. Scott is in the studio. Let's bring him in. [Music] Welcome, Scott. There he is. How's he going? What's happening? Good to have you back. Yeah, good to be back.
2:25:06
And by the way, I love Dave Tenra.
2:25:08
It's uh it's it's it's so fun to be on backtoback with him.
2:25:12
I love the I I love the story of of of recruiting and truly just fighting for the best people.
2:25:16
It's I' I've always felt the same way, which is, you know, people always ask about evaluating talent, which is of course, you know, it's it's a very deep thing, but but if anything, I think the much more valuable thing is just you kind of already know who the the really great people are.
2:25:31
And the question is like how how can you how can you build your whole whole organization around, you know, making it the right one for them to join. Yeah. Yeah.
2:25:38
The um sort of a random story, but but uh was reminded of this.
2:25:44
I think we talked about it a couple days ago, but Steve Jobs apparently called the founder, the original founder of Siri like over 20 times.
2:25:53
He called him every single day.
2:25:56
He knew that he wanted Siri to be a part of Apple and he just repeatedly called over and over and over and over and over and over.
2:26:01
Wouldn't stop until he finally said yes to uh joining.
2:26:03
So, but I think some something about it too is like I think to you you said something which is like everybody knows who the best people are.
2:26:12
But I actually think that's maybe only the case if you're also ultra No, it's easy. Hiring is easy.
2:26:17
You just look up who's won an IMO gold medal.
2:26:18
You hire all of them and then uh you just raise a bunch of money and then the rest is easy. It's super easy. No, it's it's fair. It's fair.
2:26:27
Yeah, you have to be you have to be in the right kind of circles and communities to know the best people.
2:26:32
But but I I generally think that people spend relatively too much time I think on evaluating you know people who are at the border uh versus like really just fighting for the the absolute top tier people. So yeah. Yeah.
2:26:44
It's always it's always a mistake to think like oh this is like a B player.
2:26:48
I I can make them into an A player.
2:26:51
Like there's just so much sweat there and uh the math rarely works out.
2:26:55
Usually you're better just finding someone who's a great fit for your company, your team, the skills that they have, and you can just uh turn them loose and let them run, which is always the best.
2:27:05
Anyway, uh I mean, tons of news.
2:27:07
Uh every company seems to have an AI coding agent.
2:27:10
I want to know how should I be thinking about the market.
2:27:15
I was uh I was uh on X talking to some people.
2:27:19
I had kind of mapped it out as there is consumer code that is written when I just prompt 03.
2:27:27
So I the example I gave was I wanted to know how tall a desk was in an image.
2:27:29
I uploaded the photo and 03 without me asking it to write code wound up write writing a ton of uh image processing code in Python. I didn't say write code.
2:27:42
It just wrote code for me.
2:27:42
That's kind of autonomous AI coding.
2:27:44
Now there's codeex uh which feels more proumer. I have a GitHub repo.
2:27:49
I wanted to go fix a bug for me.
2:27:51
Then there's something more in the enterprise side like hands on the keyboard all day long.
2:27:56
I need a superpowered IDE.
2:27:57
That's Windins surf cursor.
2:28:00
And then I'm seeing the top down enterprise deals from companies like Cognition.
2:28:05
So how is that the reasonable frame for thinking about the market or are there is it all just winner take all? Is there one market? Are there four?
2:28:14
How are you thinking about it and what are you seeing in the market these days? Yeah, for sure.
2:28:18
So, so high level breakdown.
2:28:20
Look, I think there's a few different product experiences.
2:28:21
And by the way, it's fun to see so many agents coming out this week.
2:28:26
Uh, but there's a few different product experiences, but I think there's primarily two markets is is how we kind of think about it.
2:28:32
And it's, you know, in one line, it's basically engineers and engineering teams versus non-engineers.
2:28:38
Uh, I I think the thing about it with DevTool, you know, totally agree with your point obviously that there's there's some differences between enterprise and self-s serve and so on.
2:28:46
Um in practice I mean you know we have both businesses for example like and a lot of these other players that you mentioned also have both businesses.
2:28:53
In practice, I think what happens with developer tooling is, you know, engineers are engineers, right?
2:28:59
And so, uh, engineers are talking to the same people, you know, the same people who are building their own projects at home are also going and, you know, shipping, uh, uh, pull requests and stuff at work.
2:29:07
And and there's kind of a lot of overlap in this that space.
2:29:11
that space. And so there's kind of the the the I'll call it the like kind of like the app builder category of like you know build me a simple app and and that's kind of like targeted towards you know non-engineers today and giving them the ability to to write software build
2:29:24
software um like you're kind of saving and then there's there's engineers which there are kind of some subcategories within but but I I I think for the most part it's all it's all connected much more so my current AGI test I've moved the I've moved the goalpost 25 times I'm
2:29:39
moving them again my current and my current AGI uh goal is if I can go to Google's new uh jewels AI agent and if I can with one prompt say build me Google reader and it does it then AGI has arrived that's my new goal to yeah it's kind of insane if you think about so so
2:29:58
one of the things you know there there was uh some research that came out and kind of people did some statistical analysis and if you just look at you know the length of time that an autonomous coding system can go uh in between every human intervention It's It's kind of insane. It's doubled like
2:30:11
It's doubled like every 3 months. Yeah.
2:30:13
And so it's like that's like 16x a year, you know, and we've been doing this for a couple years now.
2:30:18
So we're like, you know, I mean, not that long ago was like a couple seconds at a time and now you're doing hours of work, right?
2:30:22
But but it's like I just mention that because, you know, Google readers obviously it's like a really great product.
2:30:28
You have to imagine they put like a million hours into that, right?
2:30:30
And there will be a point at which it is, you know, that window a million hours of work.
2:30:33
But but yeah, we we'll get there. We'll get there. Yeah. Yeah. Exactly. Yeah.
2:30:37
We had Will Brown on the show and he was saying that uh we we we have AGI but I call it 15inute AGI and so yeah you can go and do something but if it's if it's if it's a task that would take a human a week it's not going to be oneshotable in the in the traditional sense.
2:30:52
How do you and the team think about the balance between wanting to create product experiences that somebody could just be like almost like a demo, right?
2:31:01
Like somebody was posting yesterday that they like oneshotted a calendar app and like that's like cool and and you know exciting and all that but maybe not uh massively valuable.
2:31:12
But I'm curious, do you guys have some sort of like internal framework for how you decide whether something is like really worth spending time on?
2:31:22
Is it like a truly important problem uh or or is it just is is an opportunity just an opportunity to go viral and get attention but doesn't create that enduring value. Yeah. Yeah. For sure.
2:31:35
So I mean it's in the long term our target has always been you know let's let's do real engineering work that that is you know that that that teams are actually products that teams are really building out there in the real world.
2:31:45
Um and I think the thing that's kind of interesting is is yeah because it is all you know one group of people.
2:31:50
I I think it is uh the ability to to build all these like cool things out does matter.
2:31:54
It's uh we have a distinction internally where we call it street performer Devon.
2:31:58
Basically it's the the version of Devon, you know that that that where where people are asking Devon to just do really cool stuff, you know.
2:32:05
Um and it's something that we spend some time on, but obviously the the the core by far is really just like how do we have Devon solve real issues and real projects um for your codebase. Yeah. Yeah. street performer. I like that. I do wonder.
2:32:16
Uh so yeah, it feels like there's still two frames of mind uh this consumer and enterprise even for stuff that I would call it I would call work.
2:32:26
There's certain times when like I don't want to open an IDE or I don't have time in my day as a live streamer and someone who does most of their work over email and phone um more of like the executive role.
2:32:39
Uh I I still will want an engineer to intermediate the products that I want to build.
2:32:44
And I'm wondering if there's a timeline for when that stops happening or I start instantiating whole software products.
2:32:52
This idea of like the it feels like most of the time when you hear someone say, "Oh, I used AI and I vibe coded an app in the weekend that they're still a programmer and they're still running a business and they're still working on that 40 hours a week or something like that, but they're just getting a lot more leverage to this.
2:33:09
I I feel like the really cool moment that's maybe coming is when I I I actually say I want Google Reader and it just builds it for me.
2:33:16
Um and so so in terms of shifting that enterprise experience of like real not street performance but real software into the consumer uh into the consumer space does that come top down or does that come bottom up from uh like the chat box that I'm already interacting with? Yeah.
2:33:32
And so so I I I think both forms are going to happen in the sense that you know big companies are going to look look there's there's obviously always competitive pressures.
2:33:42
There's a lot of relative things going on and and I think you know big companies that already have like you know software products that millions of people use.
2:33:49
There's going to be a lot more onus on them to just make really really great software.
2:33:52
You know it's I mean it's I I I keep track of this every day.
2:33:56
You know when when I'm doing things and I run into just bugs in software in the stuff that I'm using.
2:34:00
It's it's it still happens all the time.
2:34:02
I mean, let's let's be honest. Try and book a flight.
2:34:06
X is uh X is the the absolute worst culprit here.
2:34:09
The amount I I interact with more bugs on X than any other piece of software I use. I completely disagree.
2:34:15
I I mean, it's so much worse if you go and try and book a flight on the American Airlines app. The bar is so low. I don't know. I Yeah, go ahead. Go ahead. No, go.
2:34:25
No, I was going to say, yeah.
2:34:28
So, so that's, you know, their like product quality is going to have to get way way better there.
2:34:31
have to get way way better there. I think the other side is also going to be true which is individuals are going to be be able to build cooler and cooler apps but but yeah I mean I think the high level kind of general trend is is
2:34:42
um as we're saying you know as these tools get more and more autonomous like you want to be able to just use them for longer and longer stretches of time which is I mean which is which is pretty topical given all the agent launches that we've seen this past week. Yeah. Yeah.
2:34:53
Uh, talk to me about positioning against Google's offering.
2:34:56
My my experience from as a consumer.
2:34:59
I only interact with uh, Google right now on the AI front.
2:35:02
As a consumer, uh, obviously I've like run ads on Google from a business perspective, but in general, I see a Gemini launch and then as a consumer, it's very very hard for me to access that product that I have five different Google accounts.
2:35:15
I can never tell which one is on the Google AI premium tier. I got I finally got in. I got access to VO3.
2:35:22
It was an incredible model.
2:35:22
Feel like it felt like they were truly uh a step above and there was kind of a chat GPT style moment for for video that given that the audio comes through.
2:35:32
It's super photorealistic.
2:35:34
It's a great experience and it's and it it justified me not only paying $500 a month, but going through the UX hurdle.
2:35:40
going through the UX hurdle. Is is the developer experience on the enterprise side dealing with Google similar or are they more dialed in there because it's more important because I imagine that one of the advantages that you have as a startup in the enterprise sales space is
2:35:56
that you can do founder le sales you can take the needs of the enterprise into account much more uh different SLAs's more custom solutions there's so many more things where you can iterate even if Google is able to come out with something that uh is like wow okay they're really good at at research but on the product side maybe it's lacking. Yeah. Yeah. So, so it's you know Google Yeah. Yeah.
2:36:15
So, so it's you know Google has jewels now open AAI has codeex launch you know GitHub is doing doing more with agents as well and can I just tell you I mean it's the we have always had the view that just like everybody's going to go do this eventually you know.
2:36:29
No it'd be really bearish if everyone was like oh no Scott's like just kind of crazy.
2:36:32
I don't know what he's up to.
2:36:32
I don't think it's a good idea.
2:36:34
It's like no you should inspire a bunch of people to be like that's a good idea.
2:36:38
we're going to do it too and then you should still beat them. Yep. Yep. Yeah. Yeah.
2:36:44
And it's like parallel asynchronous coding agents, you know, has been it's been the core bet of our entire company basically for the last year and a half.
2:36:52
Um and I think it's it's really interesting to see. Yeah.
2:36:53
And so so I think there are a few factors going on.
2:36:56
Like I I do think there will be multiple players, you know, in the long term.
2:36:58
But but to your point, John, I mean, I think there's also just a lot of uh there's a lot of meaningful differentiation.
2:37:04
lot of meaningful differentiation. And I think the thing that's so interesting about the coding agent space is like it is truly you can kind of look at most spaces and and kind of call it I'll say like this is a capability space this is a product space right and to just give you a kind of a few examples I would say
2:37:20
you know it's like um image generation is like sure you know there's a lot of kind of like icing you can put on and things like that but at the end of the day you know if you just have a way better image generation model you know that is the that is the the kind of the the core like that is the meat of what you're building right um and like on
2:37:35
other spaces for example you know like customer support or something like that you know it's no one would disagree with you that the capabilities are there and a lot of it is just like what is the right interface for humans and AI to really work together and uh and get that going with coding agents it is truly both where it is you know on the one
2:37:52
hand obviously yeah I mean every every IQ point really matters you know in the system that you build and every every piece of data that it can use and analyze and work with it really matters um and then on the other hand it is you know it's the software is just messy enough and deep enough that it's it is it's it you do want to have a lot more control of what's going on. You want to
2:38:10
You want to be able to specify these details.
2:38:12
You want to be able to check in and review the work yourself and all these other things.
2:38:16
things. And so there's a lot of deep work with interface and it's uh yeah I mean in short I think these are these are the kinds of problems that we've been thinking about for the last uh for the last year you know and so it's uh it's fun to see and and you know the fun
2:38:26
thing too actually is like um I mean we're we're up 30% week over week actually with just the and it's and I think there's kind of a funny thing which is like you know people people know how to use chat bots right uh and it's like something that we've all kind of learned over the last couple years. I
2:38:42
I I think that the the agent form factor is just really different, right?
2:38:46
And and it's something that people have to learn and have to understand and and now that everybody's talking about agents, it's actually, you know, it's actually been great for us.
2:38:53
So, do you think deep research is part of that?
2:38:54
Like I feel like deep research for a lot of people is their first first uh introduction to agentic uh AI in the sense that it's like you fire it off and you let it go for 20 minutes and then you come back, right? Yeah. Yeah. Yeah. For sure.
2:39:08
I think the whole right asynchronous kind of action is is really like a big piece of it and I think it's you know we were talking about this last time but it's like agents are going to be everywhere you know pretty pretty quickly and there's no reason you shouldn't have an agent who's you know going and buying your flight so you don't have to go on the website yourself or going and you know all of these different things.
2:39:27
Um and and I think for a few reasons code has kind of been the first place where that's really taken off but but but yeah this is a total wild card so feel free to pass if you haven't dug into it but uh Ben Thompson was talking about the agentic web.
2:39:40
we're talking about agents now.
2:39:42
Uh, and he was kind of noodling through the idea that stable coins could be a relevant tool in the toolkit for the agentic web as you, you know, ask the one of the examples was plan a birthday and it's doing a lot of things on online.
2:40:00
Maybe it's buying uh birthday cards.
2:40:02
It's also looking up a cake recipe and then printing that out.
2:40:04
Well, that cake recipe website is monetizing with advertising right now.
2:40:08
stable coins in the future, you could just pay a tenth of a cent for that page visit.
2:40:12
Are you thinking about stable coins at all? Does that sound real?
2:40:19
Just if you nude on it, how do you think about this stuff? Yeah, for sure.
2:40:23
So, I mean, I think with, you know, building software engineering age, we're probably, you know, on the further end of away from it, but but with that crypto pivot. Yeah.
2:40:29
Not a lot of Yeah, I imagine that not a lot of customers are paying onetenth of a cent for Devon right now.
2:40:35
that's probably more in the said, you know, it is like an interesting thing which is like maybe one way to put it is like a lot of the norms of the internet to change, right?
2:40:45
You know, it's we have captures for example is kind of how you prevent bots and and it's it's almost like captures aren't the right like system where you actually now want your agents to be able to go and like navigate on the internet and do things for you.
2:40:54
And like a simple example of that for us is like this is again totally unrelated to software engineering, but it's like you know we give Devon our ramp card and then Devon just goes and orders groceries for us and stuff like that.
2:41:04
It's like, yeah, it's great, but but you can imagine, you know, there's a lot of these systems of like, yeah, how how should agents interact with money in the real world?
2:41:12
How should they u, you know, h how how products and and and websites and stuff have the right incentives if it's humans that are going to them anymore, but agents and humans?
2:41:22
And so, yeah, a lot a lot of questions.
2:41:23
Has Deon ever missed a memo on a ramp uh receipt or is Devon pretty good about that?
2:41:29
I think Devon's got it all down.
2:41:29
Yeah, Deon takes screenshots and stuff as well, so it's all Yeah.
2:41:32
How are you thinking about the the sort of stack, right?
2:41:37
Because I know you got a partnership with Linear.
2:41:40
Obviously, we work with Linear.
2:41:42
Uh there's idea that uh much like a member of like a team, an agent can kind of appear in a lot different places, right?
2:41:50
Uh a teammate can be in the office, they can be in a, you know, in a Zoom call, they can be in Slack, whatever.
2:41:56
Um, and how are you thinking about kind of like a aentic kind of like orchestration, right?
2:42:02
And sort of managing agents. Yeah. Yeah.
2:42:07
Linear is, by the way, they just had a launch this week.
2:42:09
They had launched this morning actually. So, so yeah. Yeah.
2:42:13
Shout out to the linear team.
2:42:13
Um, and yeah, it's it's uh it's an interesting one.
2:42:16
I mean, speaking of this kind of like product interface question, right?
2:42:19
Um you know the the work that you have to do is really like u is is you know I think one of the difference between chatbots and agents is an agent is truly like it truly needs to be like basically an employee in the same way you know it is really like you
2:42:38
as an engineer are using your team of of of junior buddies who are going to go and like implement this stuff for you right um and um and and and so like a lot of what that means is it has to be in all of these same systems, you know, in order to work really cleanly. Um, and
2:42:50
Um, and like the the the standard workflows that we have.
2:42:54
I mean, we have Slack, for example, we have uh linear and JR and so on.
2:42:57
But it is a big part of it. Yeah.
2:42:59
Last question uh actually from your post.
2:43:01
You asked for questions from people who follow you and Remy Olsen asked great one.
2:43:06
What do you see as the primary limiting factor for coding agents between context window reasoning which I imagine is just the amount of of inference uh you're delivering to reasoning tokens managing complex environments or other things.
2:43:23
Yeah, I think of those I would say I I would say managing complex environments is probably the biggest one in the sense that look if if you like speaking of 15minute AGI you know if if you give it like a a very contained coding problem where it's just like all right write one you know one file of code that solves
2:43:38
this thing it's honestly it's already insanely good you know AI is is already basically there I would say on that I I think that the areas where you see these agents or or models in general like really slip up is is kind of stuff where like you actually just need outside information to go and do what you're doing. You know, it's it's like the
2:43:58
You know, it's it's like the simple examples are, you know, if you're fixing bugs or you're implementing things like you want to be able to test the code yourself.
2:44:05
This is a big part of why connecting with all these systems is so important.
2:44:09
And so, you know, plugging into Slack or Linear, for example, makes it really easy for you to kick off your dev and runs or whatever it is.
2:44:14
But I think a big part of it too is like if you guys are actually doing your issue tracking in linear or if you guys are you know this is how you do front-end deploys and this is how you run the llinter or something like that.
2:44:23
It is obviously you know the the agent is going to be super handicapped if it doesn't have access to the same things.
2:44:30
Um and you know that the next step beyond that obviously is not accessing those things but but learning how to really intelligently work with these and use them to make decisions.
2:44:38
Any reaction to the llama 4 behemoth?
2:44:40
I I I think I saw 10 million token context window seem bigger than ever at the same time kind of going back and forth on how that project is progressing.
2:44:49
There's been kind of mixed reactions online.
2:44:51
What what's been your take away from that?
2:44:54
Yeah, there's a lot of cool projects working on longer context windows.
2:44:55
working on longer context windows. I I would just kind of give as as one point which is uh it's interesting you know if you think about human uh human human thought which is which is not always the best parallel but but is is is obviously is is the one example that we do have you know humans actually have terribly small context switches you know it's
2:45:14
like when you get like the six-digit code you know it's like like I don't know it's like pretty hard to remember what the six digits were and type those in goldfish mode yeah even with coding it's like it's like once you have to remember like the three different files that it takes you a while to spin up the context window and then you don't want to get pulled out of coding. You want to
2:45:31
You want to be in the flow, right?
2:45:34
And so the thing that humans are like really good at I would say is more like intelligent retrieve, right?
2:45:40
To your point, which is like you put this file and you're looking at it and you're like, yeah, like I remember why we did this two months ago and then you're kind of like you can kind of like replay it in your head almost.
2:45:48
That's kind of the thing that's that's that's really interesting.
2:45:51
I I don't think it's so so I I guess long story short is like I I think long context is going to help a ton.
2:45:55
Uh it's going to make things a lot better.
2:45:57
But but I think there are some kind of like fundamental things of like how do you access all of the information that you need access to to make this decision that's right in front of you right now which is like context alone doesn't solve that problem. Sure.
2:46:09
Well, thank you so much for coming on. This is fantastic chat.
2:46:12
We love to talk to you again soon. This is great.
2:46:14
Have a great rest of your day. We'll talk to you soon. All right.
2:46:17
And up next we have Joe Weisenthal, the stall war. 30% week over week. Pretty big. It's Wednesday. It's Wednesday.
2:46:24
Uh well, welcome to the studio, Joe. Bring him in. Bring him in. Bring in Joe Wis.
2:46:30
I don't want to wait anymore. Oh, welcome to the show. Thank you. How you doing? Thank you for having me. I'm really good.
2:46:38
I'm thrilled to be back here. How are you holding on?
2:46:40
What are you doing on the market?
2:46:42
The market was flat on Monday and we didn't know what to do.
2:46:46
Normally, we're either depressed and praying for something to change and then or celebrating a big update.
2:46:50
We've seen these like 5% swings and it's just been boring. It's so painful.
2:46:55
I thought like, you know, the market stopped moving and I was like, "The technology brothers will never have me back.
2:47:04
They'll never have a reason to call me again."
2:47:06
My I had I had I had one month of relevancy.
2:47:09
People were listening to my podcast because the lines were moving up and down and then they stopped and I'm like, "Okay, well, I enjoyed it while it last." Back to obscurity.
2:47:17
Yeah, we're we are having you on to We want to have you back on the show.
2:47:21
We're having you on to move the market. Yeah. Yeah.
2:47:24
This will drive the market.
2:47:24
We want to create some volatility just off of some.
2:47:30
Well, I mean, so so with the trade war, uh, we Yeah. Yeah.
2:47:33
uh, we Yeah. Yeah. I I I kind of had this like somewhat controversial take that like the the trade war was not like COVID or the or the interest rate crash because the the person who caused it could immediately undo it and Trump could roll it back which he did which he did and I was talking to Jordan
2:47:53
Schneider trying to talk about this like he was saying like no this time is different it's not going to be like the trade war from Trump won we're going to look back on this as like the beginning of the end it's terrible he was very doompilled but it feels like now if we flash forward four years. We could be
2:48:04
We could be looking back on this and being like that was a crazy but it wasn't that big a deal in like the world historical context.
2:48:11
What's your read on it now?
2:48:12
It's a really tough call because I was very sympathetic to that and as soon as he walks it back you're like you're just going to walk it back.
2:48:18
You're so stupid for everything this time.
2:48:20
All that being said, we have a laugh track now. I love that. It's so professional.
2:48:26
All that being said, a 10% baseline tariff on the rest of the world is not nothing.
2:48:30
We are now a high tariff country. Oh yeah.
2:48:34
Um there there is this pause 50% tariff or whatever it going to be is not nothing.
2:48:41
These not complete nothing and I do believe that they will change the economy in some ways that they anticipate.
2:48:48
Look when he unveiled that chart on April 2nd I thought it's like well he's basically unplugging the economy.
2:48:54
We can't function with these levels.
2:48:55
And he he sort of realized that too.
2:48:59
He said the bond market got yips.
2:49:00
He went back and then uh he probably heard stories about potential empty shelves of places like Walmart and Home Depot and stuff and then he dialed it back further.
2:49:08
We're still in a new era though of trade.
2:49:11
It might be more like Brexit where it's like, yeah, if you go to I was in London a few weeks ago, it's still the same old UK.
2:49:17
It's still pretty like rich country.
2:49:19
It's doing well, but it's they haven't grown that much.
2:49:21
It hasn't been particularly good.
2:49:23
So it might be the kind of thing where we don't notice it dramatically except maybe it's we don't grow as much as we could. Who knows though?
2:49:30
Can you I don't know if it's the weather but when I'm in when I'm in the UK it's something about the slow growth of the economy that's just you can smell it. You can smell it.
2:49:42
Well well hopefully it doesn't cover America.
2:49:44
Uh can you can you tie this to the credit downgrade? Uh how serious is that?
2:49:48
Should we care about what Moody's thinks about?
2:49:50
I don't I don't care about credit ratings.
2:49:54
I mean, I'm of the strict view that we borrow in currency.
2:49:56
We borrow in dollars and print dollars. So, we'll never default.
2:50:00
There is a you know, I would say there's two things which is one that um you know, it looks like we're going to get some sort of inflationary mix potentially between tariffs and tax cuts.
2:50:10
So I that's tends to not be very good for treasuries and to the extent you know that like policy changes in the US are now everyone understands that they can be unilaterally changed on a whim probably not great all things being all things considered for the perceived institutional stability.
2:50:26
So that's probably also not great uh at the margin for the international proclivity to invest in the United States.
2:50:34
You know the one thing that's really notable is that a lot of markets clearly have rebounded to an extraordinary degree.
2:50:40
Most notably the stock market. Yeah.
2:50:40
Um the dollar is still very big. Bitcoin high. Bitcoin alltime high. Amazing.
2:50:46
Uh the dollar is still Well, can I by the way Oh, Pomp was on earlier.
2:50:53
Can I just say a surprise guest?
2:50:54
That's Can I just say um the the first time I was on this show, one of the questions the two of you asked me was, "Will Spax ever come back?" Yeah.
2:51:03
And I said, "Well, in my lifetime, they will definitely come back."
2:51:06
Pumped it has a new spec.
2:51:08
When I said when when I said that spaxs would come back, I was thinking like 5 10 15 20 years.
2:51:13
I did not think they said Pomp doesn't have the balls to launch a spa in the year 2025 and he proved them wrong. He proved him wrong.
2:51:24
I've never had a prediction come true that fast as that one.
2:51:27
So, I appreciate Pomp vindicating uh my last appearance on TBPN. Yeah, it's great. Vindicated.
2:51:37
Um, I want to talk about like like the Middle East deals, the what was your read on on all of that?
2:51:42
It didn't seem like it was super market moving per se, but it was very big in tech and it seems like there's this new eration there.
2:51:54
It seemed like like amid all the the turmoil of the tariffs, there was like bright spot of like some really strong new alliances.
2:52:01
Obviously, there's some controversy there, but what was your read on the on the Middle East news?
2:52:05
You know, I I I've been posting about this a lot, you know, since I hit uh my mid-40s, I've been reading a lot about the Cold War and mid- 20th century history because it's what you do at my age and so forth.
2:52:16
But one of the things that I've been reading about lately is that one of the, you know, every time the US and the Soviet Union appear to have some period of Dayton, so to speak, one of the entities or one of the forces that pushed the countries apart were the sort of like human rights groups, etc.
2:52:32
uh that really uh didn't like you know various activities uh going on in the Soviet Union and that often made it difficult to have peace.
2:52:38
It's an interesting thing to think about well what if foreign policy just became a little bit more transactional.
2:52:44
We sort of accept that like there's like a very different system in this country than uh you know this other country.
2:52:51
But you know what maybe we can have stable relations if we agree to sell them cuttingedge GPUs to set up an AI.
2:52:58
It's an interesting thought and I don't have like extreme or strong views on foreign policy. It's not my wheelhouse. Yeah.
2:53:08
But it's an interesting thing to think about.
2:53:10
You know, for so long there was this view that you know like an important aspect of stable relations was some sort of common ground in governance and values and so forth.
2:53:21
And maybe we are entering this era where stable relations are who we agree to sell chips to.
2:53:29
Uh on the other side, it's an interesting thing to think about.
2:53:34
Jensen, another hundred billion for Jensen.
2:53:35
Yeah, he's been on a terror. Jensen keeps winning. Yeah, except in China.
2:53:39
Have you been tracking how many conferences he's been going to?
2:53:41
He's he's it feels like he's at a different conference every single day.
2:53:45
And yet, normally that would be very very distracting for a CEO, but every time he goes to a conference, he leaves with a hundred billion dollar deal.
2:53:52
So, it seems like it's working.
2:53:54
It's really It's really remarkable.
2:53:56
And you know again I don't know how much like I've been trying to think about this like because the way people are talking about AGI and you guys have probably a lot of guests who know way more and being out there is like you know the race to achieve AGI is something akin to like which country will develop the nuclear bomb first we must get there.
2:54:15
I I don't know it feels a little sci-fi to me.
2:54:19
I don't know how much to take that seriously.
2:54:22
Other people are better equipped to talk about that.
2:54:23
But there is this sort of like simple business storyt.
2:54:34
And so to the extent like just from a business standpoint, Jensen going around the world making all these sales potentially boxing out Huawei uh from the Middle East just from a business standpoint seems to be doing very well.
2:54:47
Yeah, we were just talking to Scott Woo and uh from Cognition.
2:54:49
They have an AI coding agent and this idea that we have AGI, but it's just 15minute AGI.
2:54:54
Like you can you can get an agent to go and do 15 minutes worth of research for you and compile a little uh a little dossier for you in deep research.
2:55:04
You don't have AGI to the extent that you can just say go create economic value for the rest of your life like you can with a human and the economic incentives just lead to a person finding their way in the economy and finding a job and buying things and consuming things.
2:55:18
I still haven't had like a great experience yet with any of the agent tools.
2:55:23
I was um someone asked me they're like do you have a list of all of the Twitter handles of guests that have appeared on OddLotss?
2:55:29
Um and I was like no but let's see if I could do that.
2:55:34
So, I went to Manis, the Chinese Manis.
2:55:36
im, and I said, "Can you put this together?" Oh, yeah. Sorry. I know.
2:55:38
Um, and then it's pronounced Manis, too, by the way. Manist.
2:55:44
And then it it it worked for like like it was like 30 minutes later.
2:55:48
I went back and checked and it said, "Oh, you're out of credits."
2:55:51
So, I did I never even got to see if it worked yet. I don't know.
2:55:54
I have yet to experience.
2:55:56
Well, the question for me is is how long would do it take, you know, like a like an intern to do that job?
2:56:02
Well, it take a really long time, but like I don't even like but you know like my experience I think we may have talked about this last time like my experience like I'm still not totally convinced that they're like reliable yet.
2:56:16
I don't know like I I I think you make a good point.
2:56:21
It's just that uh when I think of 15minute AGI I think that what's a 15-minute job?
2:56:25
It's look at the most recent OddLot show notes and find me the one Twitter handle for the guest. Yeah.
2:56:32
And now if you want to do that 200 times, that's not 15inut AGI.
2:56:34
That's that's one week AGI.
2:56:37
And you need to spawn all these different processes and orchestrate them and and fact check them and build that up.
2:56:43
And so yes, man should be able to do that, but it's not I think it's just wildly unreliable right now.
2:56:49
And so sometimes you get an output that is 10 times faster and than your best employee.
2:56:55
And then sometimes it just completely fails.
2:57:00
And we're not really equipped to like process. Okay, is the agent good?
2:57:02
Okay, well sometimes the agent is incredible and then other times it's like yesterday I wanted it to count a bunch of boxes and it like worked on it for 14 minutes and failed. I completely agree.
2:57:12
Like I'm I'm every day still.
2:57:14
And by the way, uh I've the music app I'm a songwriter. That's incredible.
2:57:21
That actually of all the AI tools that I've played with, that might actually be the one that's most incredible.
2:57:28
I've been doing this thing where I just sing a song into my phone and then I say now produce this to sound like an 80s country song and it actually works.
2:57:36
So that like some of these things just like absolutely blow me away.
2:57:38
I have a thesis for that because with w with music there's no right answer like there's no there's no oh you got one character wrong in the in the Twitter handle and it's not going to resolve right.
2:57:51
But if you get one note slightly wrong but it still sounds like music.
2:57:55
you can kind of chalk that up to the vibe or or slight differences.
2:57:57
vibe or or slight differences. And so I feel like that's why the the image generation it's less like binary or you have code where it compiles and it can work against itself and check okay does the code work or not try again try again
2:58:13
try again and so you either need to be when you're in the messy middle of like human management decision making analysis that's where AI is really struggling right now so just going back to the question of like can can AI compile all of the Twitter handles from every guest I'm guessing the answer Yes. On the other hand, if that's what AGI
2:58:30
On the other hand, if that's what AGI is, I'm not really sure it's existential whether like China the Chinese model is better than open AI model.
2:58:41
AGI is when you chat GPT to bring down the long end of the yield curve because it's really it's really getting up there.
2:58:47
And right now, if I could just go chat GPT and say, "Hey, let's do something.
2:58:50
Can you help us out here?
2:58:51
Let's do something that we need to that we need to develop right away." Yeah. Yeah.
2:58:55
So, let's let's talk about this.
2:58:55
I I know you have a project level urgency. Yeah. Yeah.
2:59:00
Uh I know you have a a podcast episode coming out tomorrow about uh about bond yields globally, what's happening.
2:59:08
Could you give people a little teaser preview on on on that and then they can go listen to the full thing?
2:59:19
Absoluter standard chartered, one of our favorite thinkers.
2:59:21
We actually never had him on the show.
2:59:23
you know, the story there's there's a there's a few moving parts to what's going on, but a big thing is this combination of countries around the world are spending more money.
2:59:34
They kind of have to because, you know, like again, we talked about US foreign policy.
2:59:38
Um, Germany doesn't know if the US is going to be the same type of ally.
2:59:43
Trump for years, even going back to his first administration, obviously encouraging the NATO countries to sort of more formally meet their commitments.
2:59:50
So you have increased uh spending.
2:59:52
At the same time, you have tariffs which are going to make business more costly, less effic um probably less efficient because you're not going to have the most optimally designed global supply chain that would exist in a world of uh sort of pure free trade.
3:00:05
So you have this combination of every country sort of having to become a little bit more self-reliant at least in some respects.
3:00:14
And you certainly see that in a lot of like tech areas, weapons, probably food even as well, energy and so forth.
3:00:21
Combined with the fact that, you know, you're having sort of less efficient supply chains, that's probably all things considered means um you know, more inflation and for uh to hit inflation targets, central banks are going to have to generally speaking keep rates up.
3:00:36
We might get rate cuts in the US in the short term.
3:00:38
That's what uh Steven Englander thinks.
3:00:40
That's certainly what the market is pricing.
3:00:43
But he thinks the window for that will be pretty short and maybe even by next year we could be talking about rate hikes again. Yeah, I know. I know. We need that AGI.
3:00:52
We need that AGI to bring down rates.
3:00:54
Uh how do you think the big tech companies have been changing their tune around how they talk about uh it used to be all about safety which you you know I think a lot of people have kind of just gotten sick of safety argument.
3:01:06
Now it's all product and all of a sudden it's we're not moving fast enough.
3:01:09
You know Siri isn't good. Yeah. It's confusing.
3:01:13
It's a little bit confusing because I mean first of all I agree the safety story feels very yesterday.
3:01:20
This is such a business story.
3:01:22
The market to for better or worse and I don't really have a view on safety but the market for better or worse seems to be dictating to the companies.
3:01:30
It's like no you got to race ahead as fast as you can.
3:01:34
You do not have an alternative to this and so forth.
3:01:36
And you even see it, you know, obviously like when Deep Seek came out and it displayed its reasoning.
3:01:43
Open AAI had to respond and it started displaying some of its reasoning models.
3:01:46
So there's the sheer competitive aspect that like you just sort of got to do what you got to do for like the product side.
3:01:52
On the other hand, there are things that the tech companies are doing that do not sort of seem to me, oh, within a year they're going to build or within a couple of years, um, you know, they're going to have these brains that are thousands of times more intelligent than the uh, than a human.
3:02:10
I saw there was something about open air.
3:02:13
I think about a social network.
3:02:14
They hired a CEO for like product and revenue.
3:02:15
There was that news today about buying a heart.
3:02:20
You probably already talked in does not these sound like ventures.
3:02:25
These sound like investments that are about making money for business and do not sound like we are building something insane in a lab that is going to change the world.
3:02:35
So it feels like there is this impulse among the big AI companies of various flavors to basically build good businesses and profitable things.
3:02:46
rather than this technology that you know is going to render human existence obsolete in some way or another. Yeah. Yeah.
3:02:54
And it feels like open AI as a lab there there could come a time where open AAI could kind of uh uh what's the right phrase for it?
3:03:06
But but effectively like OpenAI will be in a position where they can be like yeah we're a consumer internet company. Yeah.
3:03:12
internet company. Yeah. And there will be other there there will be other lab there'll be other labs that don't have the luxury of saying that and it'll be interesting if there's like this kind of divergence in I mean have to keep the AGI nar the steelman on the
3:03:27
AGI narrative is that like would we have gotten this amazing consumer technology and enterprise technology and all this value without the crazy AGI narrative that drove nonprofit investment in highly unprofitable research and it's it's kind of like I feel like everyone
3:03:46
makes the uh the comparison to nuclear but when I play back the story on nuclear it's like America incredible Oenheimer Manhattan project to develop bomb which I don't see really like a good thing like the bomb obviously it's just like it only destroys things and then it immediately becomes this this
3:04:05
arms race and this equilibrium is just that you have to build a ton of them but none of them get used they create no real economic value right um but the good ending of nuclear would have been energy too cheap to meter nuclear power everywhere. We didn't get that. And my We didn't get that.
3:04:18
And my hope is that we're in the situation where we have the panic scare around AGI as a weapon.
3:04:25
But we don't go the nuclear path where oh, you can't possibly have Chachi be dangerous.
3:04:31
Well, I I agree with you and I think if you look at like I mean it is absolutely extraordinary the rate of growth that big temp big tech companies are growing from.
3:04:42
I mean, this is what's sort of breaking the brains of everyone in the market, which is that everyone's been long big tech forever.
3:04:49
They're the biggest companies in the world, and they're still putting up some of the biggest earnings growth that you've that you can find anywhere in the market off of a gigantic base.
3:04:59
And I do think part of that story is got to be that CEOs really believe to and I don't know exactly what they believe and I don't know exactly what AGI is and they all have different definitions but they probably all believe it's like no like we're not going to slow down our spending on this stuff because we haven't achieved that thing that would that's out there that will be so special if it's achieved. Well, I got to ask Joe.
3:05:25
Would you rather would you rather own Costco at 59 earnings 59 Costco at 59 earnings or Google at 19 rotisser chickens?
3:05:37
You'll never get me to make you know you'll never financial advice.
3:05:41
You'll never get me to give fin why is every you tell me why is everyone so negative on Google these days? What's the story?
3:05:46
I don't I think it's products that are really hard to access like like they're incredible research lab but they but they haven't communicated very effectively around the product.
3:05:55
So they introduce the product and it's in a research preview.
3:05:59
They have a paper, but in order to access it like chat GPT when there's a new feature, you go to chat.
3:06:05
com and it just says use this thing and and with Google it's like there's the Gemini app and then there's Gemini smart responses and then but that's the product experience and then there's obviously like the the threat to the core business model wind up searching in ChachiBT you stop.
3:06:21
They got a kn they got a knife on the neck of the golden goose. Yeah. No, it's true.
3:06:27
Um, the Google AI products, some of them are extraordinary.
3:06:30
I've played around with some of the coding tools.
3:06:31
I've notebook is a very useful tool for research.
3:06:36
It is true that like no one would ever know how to find any of this stuff.
3:06:42
It's really hardc Well, it's in the app store now.
3:06:44
Notebook LM's in the app store.
3:06:45
It's in the It's notebook.
3:06:48
What kind of name is that?
3:06:48
Who would even know what that is?
3:06:49
It does seem I don't know.
3:06:50
What kind of name is Joe Weisenthal?
3:06:52
Yeah, but what kind of name what kind of name is like 03 Mini?
3:06:55
Yeah, they're all bad names.
3:06:58
The names The names won't The names won't matter.
3:07:00
The names don't matter, I think.
3:07:02
But yeah, there there's something about uh like Google is if they really really went full tilt into AI, they would just say Google.
3:07:11
com is our AI product, right?
3:07:14
Instead of there's a separate app, Gemini, and like you can go over here and we have an extra thing instead of being like no like we are an AI company.
3:07:22
Could you see that happening? Uh, no. No.
3:07:24
I don't I don't think the company set up for it at all.
3:07:25
But I think that on a long time it might be different if it's if it's Mark Zuckerberg and he owns 100% of the shares and he can just do whatever he wants.
3:07:32
It's private company and there's no investors that can sue.
3:07:35
Like there are different structural dynamics at Google than at a small startup.
3:07:39
I can just see it being question. Yeah. Yeah. Please.
3:07:41
Can I ask another question to you guys?
3:07:43
Do you have a clear way to articulate the benefit that Meta is getting from all of its AI spending?
3:07:52
I mean, yes, because the first like the the the the base use of AI at at at Meta, the number one like just energy cost in GPUs is ads delivery.
3:08:03
I know, but old fashioned machine learning algorithm stuff like what does that have to do with generative ads?
3:08:12
the same thing like like the the algorithm that decides what ad to serve you when you're scrolling Instagram is a big transformer now.
3:08:21
And so it's a bunch of weights just like an LLM, but instead of taking in text and putting out text, it takes in everything that you've looked at, how long you looked at it, uh what you liked about it, whether you better and better.
3:08:34
And so that's a huge training run.
3:08:34
They never talk about this because it's not sexy and it's not the new hot thing.
3:08:37
But but but you can think about like the the meta ads training run as similar to like a GPT3 GPT4 level training run.
3:08:44
And then the reels algorithm for what content to serve you, not just what ads, that's also another huge vertical.
3:08:52
And so when Zuck went to build out the reals uh algorithm farm, the server farm that would power the reels algorithm because it's billions and billions of data points, right?
3:09:02
Um, he said, "I don't want to get caught flat-footed again like I did against Tik Tok."
3:09:06
Because when Tik Tok came up and they and the whole algorithm shifted to I need to rank every piece of content across the entire network and match billions and billions of of videos with billions and billions of users.
3:09:18
He said, "I don't want to get caught flatfooted again.
3:09:21
Build me two data centers that are the same scale."
3:09:23
And then he had he had extra capacity thinking about, "Okay, I have two data centers because I just built double. Let's train an LLM."
3:09:29
And then the more and then the more tactical reason is that LLMs are going to get stuffed everywhere in every enterprise.
3:09:37
I've noticed and sometimes I just accidentally wander into an LL. Yeah.
3:09:42
But not even from the consumer perspective.
3:09:43
perspective. Like think about think about within within Facebook they have to process payments from their from their advertisers and they might want to run those through an LLM to say are we being defrauded or or or there's a
3:09:57
consumer uh there's customer support that's happening or any little internal thing even just like uh should should we uh should we censor this comment for a user who's under 18 years old right like that that is an LLM that runs in the background it's not consumer facing. Why
3:10:12
Why would they have to pay open AI for that?
3:10:16
Why would they have to pay Google for that when they just makes a lot of sense?
3:10:18
Can I say one more thing? Please.
3:10:20
I I need a um for anyone who's watching, I have a lot of dear friends.
3:10:23
Please never send me a real. I never watch them.
3:10:28
I like someone's like, "Oh, they send you a real."
3:10:30
It's like I'd rather not.
3:10:32
So, I would like to use your platform to let Yeah.
3:10:34
What's the best way to get my friends know I love them. Yeah. Send me a DM, whatever.
3:10:39
I'm just not going to wa watch your reel.
3:10:41
I'm going to I'm going to mail you a DVD of the best reels. That's fine.
3:10:44
I'm going to mail you a DVD, a Blu-ray of just all the best reels.
3:10:49
Well, what about taking it a step further?
3:10:51
Vinyl and it's us describing reels to Jeff. That'd be great. No, I love it. I would do that.
3:10:56
We would do that all day.
3:10:56
If someone sends me a DMs me a real, I'm like, do we really have to do that?
3:11:00
Imagine a bunch of colorful marbles falling down the stairs endlessly.
3:11:04
Imagine subway surfers in the background while Joe Rogan talks about weightlifting with Elon Musk.
3:11:12
Uh that's really high performance stuff, but you'll hear it on vinyl for sure. Thank you. I appreciate it.
3:11:18
Anyway, this was fantastic.
3:11:18
Uh let's make this Let's make this regular.
3:11:21
We This is such a fun conversation. Anytime.
3:11:24
I love chatting with you.
3:11:24
We're going to just put a calendar hold on.
3:11:27
Yeah, we'll throw something.
3:11:27
If you ever just If you ever if any part of you ever thinks, I don't I'm not excited to talk to John and Jordy today, then we'll cancel.
3:11:34
I'm always excited to talk. That's great. Yeah. Yeah.
3:11:36
Well, we'd love to have you back. Thank you so much. I'll be back.
3:11:38
Thank you for moving the market. This is great.
3:11:40
He moved the market on our stream numbers. We're up at 10. 8K today. Good day on X.
3:11:43
The live stream is popping. Google's up. Google's up 5%. 8% today. 5% yesterday. That's because of us. Joe said, "Why?
3:11:54
Why is everybody bearish?"
3:11:56
You just had to ask that question and the stock popped almost 3%. Just reading tea leaves.
3:11:59
Hey chat GPTt, take this transcript from Joe Weisenthal and turn it into financial advice.
3:12:06
Make it make it ultra opinionated. No.
3:12:08
Uh obviously no financial advice on this show, especially for my guest.
3:12:12
The question of like what does Zuck get from Llama?
3:12:14
We obviously did our Llama deep dive, but it's like such a history and and appreciation for open source and so many different platforms, right?
3:12:24
He just uh it would be hard to take meta seriously as yeah a hyperscaler as a hyperscaler if they didn't have any type of AI play they also get the benefit of learning from other people building on llama even if there's not some massive direct revenue benefit and and I think I think what Joe is is really like hitting on which is correct is that the LLM chatbased chatbot UI paradigm is not a perfect fit for what Meta's products are right now.
3:12:56
But as the as the UI paradigms mature, you can imagine just okay, we know that images are moving from diffusion to token based architectures.
3:13:07
So you could imagine that it maybe I don't want to constantly generate AI slop and throw that up on my story, but I could totally want to take a selfie and say, "Hey, uh, just with natural language, can you make the sky clearer?
3:13:21
It's a little gray today.
3:13:21
Can you just make it blue back there?
3:13:23
So, it's a hedge against as as as AI companions and just people develop relationships with effectively the models.
3:13:34
That is a threat to Meta's core business of connecting regular people that have conversations in the app that that bring people back to the app so you know they can serve more ads. So, yeah.
3:13:46
Anyways, today was a very fun show. Very fun show.
3:13:49
We got to go take some photos.
3:13:51
We had range from Eric Prince to Joe Weisenthal.
3:13:52
We really covered everyone.
3:13:55
I got I got some text messages.
3:13:55
Someone said, "Uh, great lineup today." I'll read this up. Where is this? Uh, great lineup today. Blackwater.
3:14:02
Blackwater for software engineers. Cognition.
3:14:06
Blackwater for Sam Alman's haters. Open AI.
3:14:10
Uh, lots of Blackwater folks.
3:14:10
Uh, the Blackwater of podcasting. Joe Wisenthal. Yeah. Or David Senra.
3:14:15
I think David Center is more blackwater coded when it comes to how ruthless efficiency he pods.
3:14:21
Anyway, thank you for listening. Uh, this is fantastic.
3:14:26
Uh, switch your business to ramp. com.
3:14:28
Check out Poly Market talking about using ramp ramp.
3:14:30
I want to get uh I'm going to get work on getting a Poly Market set up on PM Spack. Oh yeah.
3:14:36
Like does it actually does it happen? When does it happen? When does it go out?
3:14:42
Like the IPO markets that we've been having a lot of fun with.
3:14:44
And then as the rumors start circulating, who's the target? This is great stuff.
3:14:48
I'm gonna be we're gonna hit up all our boys, all our group chats.
3:14:52
Hey, who's got a profitable financial services company that they want to take public?
3:14:56
I want to I want to highlight one timeline post before we get out of here.
3:14:59
Leonard Heim, he's coming on the show uh next week.
3:15:01
Uh Jordan from China Talk recommended him.
3:15:06
He talks about chips and AI.
3:15:06
And I love this because there's a there's a screenshot that he's quote tweeting from, I believe, the Financial Times.
3:15:12
It says, "We still don't know how much a energy AI consumes.
3:15:16
Companies must give us the chance to understand the environmental impact of the tech we use."
3:15:22
And he says, "Yes, we do. It's 21 gawatt.
3:15:25
You count all the AI chips produced, factor in that they're running most of the time, add some overhead, and you got your answer.
3:15:30
It's a lot, and we'll only get more, but you know what? Probably worth it."
3:15:34
And I love I love that like there's so many of these opinion pieces that are like, "We could never know. We could never know."
3:15:39
And he's like, "No, you can just add it up.
3:15:42
It's pretty easy to understand how much energy it's producing.
3:15:43
So, uh, I saw this post after I already invited him, but I'm I'm super excited to talk to him next week because, uh, clearly banger, banger. Anyway, it take Smith.
3:15:51
Leave us five stars on Apple Podcast, Spotify has been really explicit about this.
3:15:56
If you like watching the show, our jobs are on the line. Our jobs on the line. They will recast us.
3:16:03
They will cast someone else.
3:16:05
Ben is making menace giving us a menacing look right now.
3:16:08
His his his thumb is over the purchase button on an Audi R8.
3:16:11
And he can only do it if you leave us five stars.
3:16:14
If you don't leave us five stars, we're never going to get there.