0:00
FROTHY BLACK MUSTACHE. I TYPE one careful line.
Search this video
0:01
Then I TAKE A MEASURED SPLASH.
0:03
ONE PINT, two lines, [music] a tidy little spree.
0:05
I'm patching up the script with a grin and a knee. The pipe start grinning. The whole room's in.
0:10
I'm still a proper programmer with a Guinness on my chin.
0:14
[music] Codex and Guinness. Codex and Guinness.
0:15
One prop, one point, and I'm feeling all right. Codex and Guinness. Codex and Guinness. Send me the next task.
0:24
[music] I'll toast to the night.
0:32
Now the second pine arrives and the comments [music] get bold.
0:33
I'm refactoring feelings in a barrel of gold.
0:37
I ask for a helper, get a fiddle in a flame.
0:40
I WRITE, MAKE IT CLEANER, THEN FORGET WHAT I NAME.
0:42
TWO PINTS, THREE PROMPTS.
0:44
The cursor [music] starts to dance.
0:45
I'm building a whole in a half drunk trance.
0:48
The base drum stops, the bag pipes bite.
0:50
[music] I'm still in the kingdom, but I'm losing the light. Codex and Guinness. Codex and Guinness.
0:55
One prompt, one point, [music] and I'm feeling all right. Codex and Guinness. Codex and Guinness.
1:01
Send [music] me the next task. I'll toast to the night.
1:17
Then I slip past the Bmer Peak and the ceiling spins.
1:19
I'M [music] SLAMMING DOWN POINTS LIKE A man with no wind.
1:21
I don't just type prompts, I hurl them like stones.
1:26
Make it do [music] wizard stuff. Now add more. Three points. Four points. The wisdom runs thin.
1:36
I'm naming every variable of [music] the cousins and kin. The pipes go wild. The crowd goes mad.
1:40
I'm one bad decision from glorious aside. Get it. Codex and Guinness.
1:51
One prompt, [music] one point, and I'm feeling all right. Codex and Guinness. Codex and Guinness.
1:58
Send me the [music] next task. I'll toast to the night. I'M NOT SIPPING NOW. I'M IN FULL ATTACK.
2:04
I'M CHUGGING LIKE A CHAMPION WITH A TILTED CODEC STACK. The keyboard's a plank.
2:08
The pub is a [music] sea.
2:12
I'm firing off props like fix everything for [music] me. Codex and Guinness.
2:24
Codex and [music] Guinness.
2:26
I'm past the peak and I'm roaring tonight. Codex and Guinness. Codex and Guinness.
2:31
Chuging then prompted [music] to roar and is white. One more pint. One more line.
2:38
[music] One more reckless brilliance and get us and get us. Ain't no pace left now.
2:44
Just fire and swing [music] and shine.
2:56
1 2 3 4 I OPEN UP CODEEX. ONE SMALL sip of stout.
3:11
A little BIT OF VIBE CODING JUST TO SORT it out.
3:14
I got A CLEAN NEW PROMPT AND A FROTHY BLACK MUSTACHE. I TYPE ONE careful line.
3:18
And then I TAKE A MEASURED SPLASH.
3:20
ONE PINT, TWO LINES, [music] a tidy little spree.
3:22
I'm patching up the script with a grin and a knee.
3:24
The pipe starts grinning. The whole ruling's in.
3:29
I'm still a proper programmer with a Guinness on my chin. Codex and Guinness. Codex and Guinness. One, one [music] point.
3:35
And I'm feeling all right. Codex and Guinness. Codex and Guinness.
3:37
Send me the next [music] task.
3:40
I'll toast through the night.
3:48
Now the second pine arrives and the comments get bold.
3:50
[music] I'm refactoring feelings in a barrel of gold.
3:54
I ask for a helper, get a fiddle in a flame.
3:56
I WRITE MAKE IT CLEANER THEN FORGET WHAT I NAME.
3:58
TWO PIPES, THREE PROMPTS.
4:01
The [music] cursor starts to dance.
4:02
I'm building a whole in a half drunk trance.
4:05
The base drum stomps, the back pipes bite.
4:07
[music] I'm still in the kingdom, but I'm losing THE LIGHT. CODEX AND GUINNESS. Codeex and Guinness.
4:14
One from [music] one point and I'm feeling ALL RIGHT. CODEX AND GUINNESS. Codex and Guinness.
4:18
Send me the [music] next task. I'll toast to the night.
4:34
>> You're watching TVN Tuesday, July 28th, 2026.
4:37
We are live from the 100 Ultra Dome, the temple of technology, not the fortress of finance, the capital of capital.
4:47
We've been having a lot of fun with Suno.
4:49
Hope you have been enjoying it, too.
4:52
I'm sure we'll have a new one available soon.
4:55
But first, let me tell you about ramp. com. Time is money. Save both easy.
4:59
Use corporate cards, bill pay, accounting, and a whole lot more all in one place. Um, bunch of news today.
5:04
Jord's laughing, laughing, laughing.
5:07
All right, I think we get we [laughter] can pop that.
5:10
>> That's enough of that.
5:12
>> Yeah, nothing like a couple pints of Guinness.
5:14
Some prompt >> engineering, some vibe coding going on.
5:19
Uh well, uh Anthropics responded.
5:19
We're going to go through that proposal, the the the facts and the proposal for what ne what happens next in the open uh open model debate over whether or not they should be banned, restricted, tested, limited in some ways, sued.
5:34
There's a whole bunch of different possible outcomes.
5:38
Uh but we'll take you all through it and we have tim joining from first adopter uh at 11:30.
5:42
But first we are going to talk about the hiring market because the Wall Street Journal has a very interesting report that large the Wall Street Journal is reporting that large companies are beginning to >> they have a large white pill.
5:58
>> Yes, it is a large white pill >> has hit the front page of the journal. >> Yes.
6:01
And I think people have been going back and forth on this.
6:03
This is a story that's that's just getting digested by the tech folks, like the actual AI lab leaders who had predicted crazy job losses and are now not really seeing that.
6:16
They're seeing productivity boosts and different uh diffusion taking time in certain places and there's new capabilities, but it's not exactly a drop-in replacement for a co-orker, at least in in most scenarios.
6:28
And that's what the Wall Street Journal is reporting.
6:32
So let me set the table and then uh we can debate it a little bit.
6:36
First I'm going to tell you about console.
6:37
Console builds AI agents that automate 70% of IT HR and finance support giving employees instant resolution for access requests and password resets.
6:44
So uh after roughly a year of cautious hiring, companies across technology, transportation, defense, and other industries now say they need more employees to work alongside AI systems.
6:56
Total victory for both humans and AI. we're working together. Peace is possible.
7:01
Uh it's an example of Je of Jevans paradox. Jeban's paradox.
7:08
When a technology makes something more efficient, demand often rises enough.
7:11
The total use and the need for people actually increases.
7:14
For roughly the past year, many companies pointed to AI while announcing layoffs.
7:20
This was a huge thorn in your side.
7:20
You I think you hated this more than anyone else.
7:24
Um and you were right to because it did seem like it was just PR spin, etc.
7:30
Yeah, it was it was a way for CEOs and management teams to save their own ass instead of saying, you know, hey, we we overhired or the business isn't doing as well as as as we would like and we need to sort of >> uh basically [clears throat] settle down for a second and uh get our mojo back.
7:48
Obviously, no one wants to say that, but I I think uh one of my favorite posts was the uh and and obviously these circumstances are never great, but the the new CEO of Xbox came out and just was like very honest about the situation. >> Yep.
8:04
>> And I think that uh more of that is necessary. >> Yeah.
8:08
Also, there's a lot of firms where they once they get to 10,000 20,000 employees, they might say, "Look, 20,000 might be the right number, but the bottom thousand people are not performing.
8:21
We would like to lay them off and then bring in a new thousand people that are better fit for the company and the current trajectory that we're on, the current skills that we need.
8:29
Maybe we need more sales people."
8:31
>> And those bottom people might be top 10% at another company. Exactly. Yeah.
8:35
So uh the narrative appears to be shifting.
8:37
Uh companies like CSX, Alphabet, Service Now, Snap-on, and consulting giant Booze Allen Hamilton have all recently signaled plans to expand hiring particularly in areas where employees can use AI to become more productive.
8:51
Uh we have Benzewig from Rellio Labs coming on at 1210 uh to talk about the difference in the AIdriven hiring market.
9:01
some very interesting data about how AI enabled firms are hiring faster than those that aren't adopting AI.
9:06
But at the same time, there's a bunch of weird dynamics in the labor market where there's way more job postings than actual hirings.
9:14
And so that can look like there's a fall-off and it's harder to get a job, but that might just be because everyone's slopping it up in the job postings and everyone's like, "Put up a job posting for everything because I'd love if somebody if some insane sales guy walks in the door, we might have a position.
9:30
So, let's keep it >> used to be somewhat of a flex if a company was like, yeah, we put up a role and we got 2,000 applicants. Yeah.
9:34
You know, it's like, well, >> that's true.
9:38
And then also, it's a little bit of a sign of like, oh wow, they have a hundred openings.
9:42
Like, they must be like growing so fast, you know?
9:46
So, but if it's just a prompt to say, oh yeah, put up like look at my organizational design and and add five roles for everyone because why not?
9:52
Why not see who comes by?
9:55
You know, we're not we don't necessarily have to interview these people.
9:59
Uh so weird weird dynamics, but we'll dig into it.
10:02
So uh meanwhile, the latest weekly US jobless claims fell to one of the lowest levels in decades, underscoring the resilience of the labor market.
10:08
The shift also reflects a more realistic understanding of AI's capabilities.
10:15
Sarah Franklin, CEO of HR platform Lattis, says many companies initially assumed AI agents could replace entry-level workers, but are now recognizing that human human employees remain essential.
10:25
Just because you have coding agents doesn't mean you're not hiring engineers, she said, adding that Lattis is seeing renewed hiring among many of its customers, including for junior roles.
10:36
Uh Robert Half uh CEO uh M Kenneth M.
10:41
Keith Wadd said AI's effect on employment has been more benign than some have feared, adding that hiring demand continues to improve and market conditions are increasingly more supportive of business.
10:51
And so I do think there was a little bit of like a successful scop with the with the AI is going to be able to do everything where I do think there are some firms that were like yeah maybe we shouldn't hire or because like what if we get it wrong and we hire a bunch of people and then AI really does catch up and we don't need those people. That's silly.
11:07
We shouldn't go through that like whipsaw effect.
11:10
Uh and so people are going back and forth on that. Bryce Roberts.
11:13
Yeah, it's interesting at least in at least in our organization which is unique and and uh very niche and there's not that many organizations that are running you know a niche technology uh daily show. >> Yeah.
11:30
>> I feel like a lot of what the the value that we get out of AI would have historically been done by not super expert level freelancers, right?
11:39
these sort of like upwork style tasks that you would do historically like an idea for a funny song, right?
11:46
I've paid to get a funny song made probably a decade ago online, right?
11:51
As just like a [clears throat] joke and now you can just go to Sununo and and make something like that.
11:56
Um whereas uh and and then there's other things like you know make a funny website, right?
12:01
I historically would maybe work with freelancers.
12:05
>> So, you're saying that I should I should take down the five open roles I have for Celtic punk uh session musicians? >> Not yet.
12:15
>> Cuz I was going to hire five Celtic punk session musicians to constantly record Dropkick Murphy's covers for us. Yes.
12:22
Every day >> and then perform that you shouldn't do that.
12:25
I'm actually closer than ever to hiring a full-time Celtic punk band to to play music to recreate songs. >> Yes.
12:35
I'm I'm closer than ever uh to doing that where that was not even on the road map uh a few years ago. Yeah. I don't know.
12:41
It's it's a good point there. Yeah.
12:42
There there's a lot of things that you uh are doing that you would never do with a full-time employee. Yeah.
12:48
uh that just sort of like fills the cracks and allows you to do more different things in your organization, but the core stuff is still like you want a person that's responsible and then you want them using AI. I don't know. Yeah.
12:57
Um the Wall Street Journal uh breaks it all down, but we went through most of that.
13:02
So Bryce Roberts, he's he's taking the other side of this.
13:05
He says he shares a screenshot of a text message says, "We honestly aren't hiring a ton right now.
13:11
AI backfilling most roles."
13:11
Backfilling is that specifically does that specifically refer to when someone leaves the company, you backfill them with AI?
13:20
So you say, "Oh, someone quit.
13:23
Let's see if like if there's Steve and Jim on two different on one team and Steve quits, you say, "Hey, Jim, can you just instead of hiring another person just do twice as much work with AI?"
13:30
Is that what this person's articulating?
13:33
I mean, obviously there's some companies that are like, "Yeah, we're not hiring anyone.
13:38
We're going for the one person 1 billion dollar company."
13:40
Like, I'm not going to hire anyone.
13:41
I'm just going to use >> Yeah, but that's rare.
13:43
Usually, usually when your business is ripping, you're like, I can't hire great people fast enough. >> Yeah.
13:49
>> And sometimes you actually sometimes you actually don't have time to >> invest into various hiring processes.
13:55
But yeah, >> uh yeah, I would read into this text, the company's just probably not like ripping.
14:02
>> That's my that's my takeaway.
14:03
>> Well, Bryce Roberts says, "Rip new grads."
14:05
Matthew Prince over Cloudflare takes the other side.
14:07
He says, "Wrong strategy to stop hiring new grads."
14:09
the right strategy, hire them and insert them into legacy teams to help them better adopt AI.
14:15
>> And Cloudflare, of course, hired 1,000 >> something.
14:19
It was a It was a crazy number, wasn't it, up there in like almost a thousand, >> four digits. >> That's crazy.
14:24
Anyway, let me tell you about the New York Stock Exchange.
14:26
Want to change the world?
14:27
Raise capital at the New York Stock Exchange.
14:29
Um, pulling a crazy rare business card. I haven't seen this.
14:35
I Oh, I I think I know where they're going with this, but let's play the latest Good Work uh real.
14:39
We're just watching reals now.
14:44
>> This is one we got is a Bernie Maid off. >> Pretty good.
14:51
>> That's from the 80s, too. That's good. >> Yeah. Yeah.
14:53
I've I've seen a few of these around before.
14:54
Up next, >> solid Sam Bankman Freed here. >> That's nice. >> Really nice. >> That's really nice.
15:03
>> I like that they actually printed these.
15:05
I think he made for balsa wood. >> This is balsa wood.
15:10
>> The acting is so >> Wow. >> Wait, John.
15:15
>> This is a This is a vintage Zuckerberg. >> 05.
15:18
>> This is a vintage05 Zuckerberg.
15:18
Let's just check the back really quick. >> There it is.
15:24
That is a patch from his Fruit of the Loom boxer briefs.
15:26
You can tell by the smell. >> Is that real?
15:30
What is that referring to?
15:32
This is on >> it athlete, you know, training card.
15:35
Don't put a piece of the jersey. >> Piece of jersey. >> Sleeve this one. >> Yeah. Yeah. >> All right. Up next. >> Sleeve it. >> Okay. >> Nice. >> Elizabeth Holmes. We do have two. I believe we have two.
15:45
But a triple Holmes is what every good collector has >> in their arsenal. All right. One last card. >> One. One card left. Three. >> Two. One. >> Oh my god. [laughter] Oh my god. Oh my god. Oh my god. Oh my god. >> All right. Turn it off. >> Very funny. Very funny.
16:03
>> What it is funny how the the like business comedy cannon has really solidified around like Elizabeth Holmes after uh Sam Banken Freed, Mark Zuckerberg.
16:13
There's like a few names.
16:14
>> I'm surprised they didn't have an Adam Newman rookie card in there.
16:17
>> I don't know if Adam Newman is like a big enough name relative to >> That's true.
16:22
>> Sam Bank and Elizabeth Holmes.
16:22
Uh it's just interesting like the different the different names that have broken out that you can do a comedy sketch that's like you know it goes as big as uh as good work does because they get you know I think millions and millions of views on their stuff.
16:34
>> All right, pull up this image from Manhattan this morning. We got sent this.
16:40
>> We've been doing on the ground reporting >> from one of our on the ground reporters in Manhattan.
16:43
There's a company called Black Sheep >> that that got 20 trucks and they are just driving them around Google's Manhattan office. Yes.
16:54
>> Saying, "Shame on you, Google. Return our $80,000." >> We had to dig in. We got very curious. >> I had no idea.
17:01
They make they make sunglasses.
17:06
>> They make $8 sunglasses that beat $350 sunglasses in an NBC lab test. Okay.
17:10
Are you wearing [clears throat] black sheep today? What you wear? >> I wish. I wish.
17:16
So, uh, Black Sheep makes direct to factory. >> Okay.
17:22
Factory direct prescription eyewear. Stop paying >> No.
17:25
This is from their own website.
17:28
They're saying direct to factory optical disruptor, which mean [laughter] this is from their website. Uh, on black sheep. >> I'm on black sheep. io as well.
17:37
It says factory direct direct to factory.
17:39
Look, I want to send some eyewear to a factory. >> Direct to factory.
17:45
>> I'll be sending it to them. >> Direct to factory. >> Yeah.
17:49
>> Uh so this company >> that is interesting >> is fascinating.
17:52
They say direct to factory optical disruptor black sheep launches 25 truck gorilla campaign against Google in Manhattan. Okay.
18:00
And then they're they're sort of like narrating their own gorilla campaign.
18:04
A fleet of 25 minimalist LED billboard trucks surrounds Google's Chelsea headquarters after the tech giant weaponized an organic search glitch to pocket nearly $80,000 in ad spend following Black Sheep's viral NBC Today Show debut.
18:21
[laughter] 25 LED trucks deployed.
18:25
$77,000 drained in 30 hours.
18:29
And then they're and then they're just continuing to market their own product.
18:34
Very interesting strategy here.
18:34
I think every marketer has had the experience of of having a campaign go haywire. Yeah.
18:42
>> Very fascinating to take to take this route.
18:46
Let's see how it works for them.
18:49
Uh, if I were if I were Google, I would say you can have your $80,000 back, but you can never advertise on Google again because I just don't know how.
19:01
>> I don't think Google would ban them permanently for this. This is ridiculous.
19:04
But it, you know, they're just going to they're just going to be like any other like as a self-s served platform.
19:10
>> But is it a good campaign?
19:11
>> But what what actually happened?
19:11
So they say how it unfolded.
19:13
NBC Today Show segment airs.
19:15
Uh, they test the retail subscription against Black Sheep's factory direct pair.
19:19
National search traffic spikes.
19:22
Hundreds of Americans search Black Sheep because they're seeing it on TV.
19:25
The organic listing breaks.
19:28
Google search engine redirected organic brand traffic to a dead-end third-party 404 error page.
19:33
And so, with the organic route broken, users were funneled into Google's paid listings. So what is their claim?
19:45
How is Google responsible for this exactly?
19:48
Sounds like user error >> because I mean you do you do have some control over your Google search results based on the web master tools you can index certain things and then also if you're noticing a 404 page you could like redirect it quickly but again if this is happening all very fast they can use your >> diffult uh but I mean it is interesting because they're probably going to get more than $77,000 worth of organic just from this.
20:12
I mean I didn't see the original campaign and I'm seeing this because this is hilarious.
20:16
But this is like a is this they they shared an AI image with tons of these like shame on you trucks. But those are real. >> These are real. >> Yes.
20:27
And are those minimalist or maxim?
20:30
>> Those seem maximalist to me.
20:32
>> But maybe they're minimalist.
20:33
>> Minimalist, I guess, in the in the display of the uh in the way they actually are leveraging the space on the truck. Black and white.
20:41
>> Truly underrated surface area for stunts and advertising.
20:44
like like th this message is sort of like squabbling with Google over this like sort of odd scenario.
20:50
But you can imagine someone using this for something much cooler and much more positive and not like this uh you know sort of unfortunate situation for them where they're dealing with the you know fallout of a Google error.
21:03
>> Well, we want to interview the truck drivers.
21:06
>> So if you're driving a black sheep truck around Manhattan today, reach out >> for sure.
21:11
>> Show Nick, >> make it happen.
21:12
Well, let me tell you about Shopify.
21:14
Shopify is the commerce platform that grows with your business and lets you sell in seconds online, in store, on mobile, on social, on marketplaces, and now with AI agents.
21:24
Uh, Ilia said, "Straight shot to SSI, so they better not be gearing up to release a work agent called France."
21:31
France would be a very good name for a for an AI agent. I like that.
21:34
Uh, I do wonder what they're going to be releasing.
21:38
Has has the SSI going to release. Is that complete rumor?
21:42
Because all all I said they would scale their research scale their research.
21:47
So that just means they've done a bunch of research.
21:50
They have some sort of architecture that they like some sort of flywheel and they're going to like use more compute and so that's why they're raising money.
21:56
I don't think they said like and we're going to release it publicly. >> Yeah.
22:01
>> But everyone's thinking like probably still LLM or something different. No one really knows. >> Yeah.
22:07
I mean I think still broadly like generally god like next level.
22:08
There are levels to vague posting when you live a vague life just like your entire life is vagory.
22:18
Anyway, uh in other news, recursive super intelligence signs a $410 compute deal with Amazon.
22:26
So funny and it's in the tech crunch.
22:29
It's in the it's in the the header, too.
22:31
Of course, that is a typo.
22:34
Uh, it says recursive super intelligence signs 410 million million dollar compute deal with Amazon.
22:42
Congratulations to recursive super intelligence throwing safety out the window.
22:45
That should be the that should be the uh the tagline because there's already safe super intelligence, but we're just doing recursive super intelligence over here.
22:52
Um, but of course the company is doing very well.
22:56
Um, they emerged from stealth in May with 650 million in funding focused on building open-ended self-improving systems and potentially compute intensive approach to AI research.
23:06
This multi-year deal is meant to provide flexibility as the company looks to scale up those systems.
23:11
Uh, Recursive's $410 million outlay represents the bulk of the company's fundraising to date.
23:19
But on a call with Techrunch, uh, with >> Hey, hey, they still have a couple hundred million left over.
23:23
Founder and CEO Richard Socher emphasized that he expected it to be the first of many such deals.
23:29
So is this I feel like normally when you see a like a compute deal signed, it's always like more complicated than just like we're buying this expensive thing.
23:38
It's usually like we're we're paying that.
23:40
I'm like we we used to be so like anti- circular deal that now I just have come to I've been so normalized by them that I expect them every time.
23:50
I'm like wait this is just there's no circularity here.
23:52
I would have expected changing hand. >> Yeah.
23:55
Like Amazon's investing in you and you're buying tranium and racking it in AWS and doing new campus and they're investing in this and that and you're investing in them.
24:02
Uh instead it just seems like it's a pretty vanilla deal.
24:05
It's like they're just buying a lot of compute from Amazon. Great. >> Seems like it.
24:09
>> Well, good luck to them.
24:09
Very excited for what they're launching. Um >> yeah.
24:13
Uh Jason, [clears throat] VP of startups and VC at AWS says part of the agreement is that we're going to co-develop him for a purpose built for these types of companies.
24:20
So fingers crossed, but it seems like we could get some some circularity.
24:24
[laughter] >> Yeah, let's hope so.
24:30
Uh, let me tell you about Railway.
24:33
Railway is the all-in-one intelligent cloud provider.
24:35
Use your favorite agent to deploy web apps, services, and more while Railway automatically takes care of scaling, monitoring, and security.
24:46
Fingers [laughter] crossed.
24:49
Well, well, here's here's a deal that's somewhat circular.
24:52
We got Nvidia revealed as a tenant for a $50 billion data center that will use its chips.
24:58
So, they're the tenant of the data center that uses its chips.
25:02
We'll talk to Tate Kim about this.
25:04
Uh, CEO Jensen Wong deploys balance sheet to back stop growth of AI computing market.
25:09
Nvidia has signed leases worth up to $50 billion for a massive Texas data center. That's very very big.
25:17
That's very very big for a single site.
25:19
Uh a previously undisclosed commitment that shines a new spotlight on the chips on the chip group's growing role in financing AI.
25:28
The nearly $5 trillion company is leasing the entire gig 1 gigawatt facility that developer Hut 8 is building which will house hundreds of thousands of Nvidia graphics processing units.
25:37
So you have to imagine that once they have these they serve something or wind up selling them.
25:41
Like these things change hands so many times.
25:43
There's a lot of different ways that this could play out ultimately.
25:46
Um, but in the uh uh can we pull up the Nvidia chart? Yeah, there we go.
25:51
Nvidia, >> big candle today. Up 3% 5. 17 trillion.
25:58
Let's take a look at Apple >> 4. 99. They crossed five today.
26:01
Uh they're down a little bit since they since they beat breached that, but they're neck and neck.
26:09
Uh Google is sitting at >> Yeah.
26:13
>> Apple running the do nothing win strategy. Yeah.
26:15
Jensen doing thousands of deals.
26:20
>> They didn't even sign the open letter.
26:22
Uh there are three companies that still I I believe still haven't uh signed the open letter.
26:27
And >> only three companies on the entire [laughter] surface of the earth. >> No.
26:32
Uh there's three major companies that are uh how do I actually get that to go away? I don't know.
26:36
You can keep looking at alphabet but there are three there are three major companies that haven't signed that Nvidia open letter about uh banning open source uh and or not banning open source uh and it's Amazon Al uh Amazon, Apple and Anthropic.
26:52
Anthropic put out a uh uh a post yesterday very very clear response sort of outlining their view on open source their stance.
27:01
uh we should go through it.
27:03
But um the interesting thing that uh Ben Thompson and uh uh was talking about today was uh the fact that Apple and Amazon haven't signed and they both have like very physical elements in the world in the sense that they're not uh they're sort of unsloppable.
27:20
Like you you can't vibe code an Amazon warehouse, you can't vibe code an iPhone.
27:24
Uh there are threats to those businesses of course and of course Apple should benefit from open- source and so should Amazon because they'll be able to serve open models across AWS.
27:34
But it's just potentially interesting.
27:37
I think the Apple standing it standing back is more just like look we're not jumping on with this crazy open letter that everyone is signing like we just have our own brand.
27:45
We're thinking different. We're doing >> well.
27:46
And based on other Apple AI timelines, I would expect them to sign it in maybe a year or two [laughter] >> potentially if they just sign it in 2028. Just like robots. >> WWDC 2020. >> We're ready now.
28:02
>> We're signing the open letter.
28:04
>> These glasses have have changed you.
28:06
They turned you a new beast.
28:06
Uh let me let me run through the the three anthropic proposals because uh it's an important response.
28:12
So Dario Amade and the Robic CEO uh responded directly to that letter summarizing uh supporting openw weight models uh that circulated over the weekend.
28:20
So uh to summarize his uh he makes three claims just to sort of clarify things that I think are important.
28:26
One he says anthropic has never advocated for a ban on openweight models.
28:30
Now that that's a blanket ban.
28:33
There's obviously like defining what a ban is and what an openweight model, what a distilled model, what a foreign model is. These things all matter.
28:39
But he has he has come out and said look we never advocated for a total ban on openweight models.
28:45
Uh two he says uh undergirling all of this is the US must beat authoritarian governments in the AI race.
28:52
He points to China but he identifies any authoritarian government.
28:56
If they get really powerful AI they'll come over here and steamroll us and you won't be free to do whatever you want to do in America.
29:01
Uh three powerful AI models may be misused to carry out cyber attacks or biological attacks.
29:06
there are there are risks to having really really powerful uh AI opensource systems just running around.
29:14
So he's worried about those three things clarifying those three points.
29:17
But he makes three uh recommended actions.
29:19
He makes three proposals.
29:21
Uh first he says uh let's continue to sanction chips.
29:23
Let's not sell chips to China.
29:26
He says we should not sell powerful chips or chipmaking equipment to China.
29:30
So this has been debated for years like going back to the Biden chip uh controls.
29:33
um everyone knows every different angle on this uh the basics.
29:39
the basics. I mean there is a pretty good argument for chip controls uh even on purely geoeconomic competitive grounds like even if you don't believe in the the the risk of authoritarian governments having powerful AI even if you just think it's like you know fancy autocomplete it's like well it's the engine of our economy and if you can
29:56
slow down arrival economy that's beneficial to you right um and so and there also seems to be basically unlimited demand for chips in America so uh by restricting sales to China that shouldn't actually hurt American chip companies all that Uh but yes, >> well they just like the their argument would be we [clears throat] fully lose the Chinese >> market. >> Yeah. >> Yeah.
30:18
>> Which is the >> second largest Yeah.
30:20
>> computing market in the world. No. >> Right.
30:22
So so I think um but but you're the counterpoint to that is you were going to lose it anyways. >> Yeah.
30:28
Um and a lot of that it stems from the fact that chi China has been building an indigenous chip supply for decades.
30:34
we've talked about going back to uh a whole bunch of their, you know, state-led, state funded chip uh and fab processes.
30:44
They've always been uh a few years behind.
30:46
Uh and so maintaining that gap, all else equal, is an advantage for the United States.
30:51
Uh the second point Dario makes is he says we should crack down on industrialcale distillation operations.
30:57
Uh this seems totally reasonable.
30:59
customer uh companies can set their terms of service and they have a right to maintain intellectual property uh with proper legal consequences for violations.
31:07
Uh Anthropic's been fighting distillation attacks, but according to them, it's not that effective.
31:12
Daario proposes policy interventions to deter this behavior.
31:18
And this is where I'm still not clear on what where that goes next.
31:20
Like what is the correct policy intervention?
31:24
There's a like policy intervention is a very very broad thing.
31:27
It can mean anything from like a tax, a tariff, a fine, uh a sternly worded letter, uh not getting invited to a golf tournament.
31:35
Like there's so many different things that policy like covers these days, right?
31:37
Um where where does this actually go?
31:41
He says he doesn't want a blanket ban on openweight models, but it does seem like one possible policy intervention would be to sort of like ban restrict or pressure open weights models that can be reasonably shown to have been distilled.
31:55
So if there's someone who's just a perfect distillation, it just gets it just doesn't quite feel right.
31:59
It's hard to quantify these things.
32:02
We don't have a binary where you run some sort of algorithm and you say, "Yes, this was distilled."
32:07
Because you can distill, you know, half on Opus 5 and then throw in a little GPT 5.
32:12
6 and then mix in some some misdraw and like just be distilling from all over the place.
32:17
Fine-tune stuff, change the flavor, change the RL environment.
32:21
There's so many different pieces of it.
32:23
And Tyler, you were making a point about uh Tinker or >> Yeah.
32:27
So, so the the Inkling model from thing machines like it used some synthetic synthetic data that was created with uh I think Kimmy K2. 5. >> Yes.
32:37
>> So like I does that count as like distillation like probably not when people usually talk about it but like it definitely benefited from Chinese open source models. >> Yeah.
32:46
So I wouldn't call that downstream of >> Yeah.
32:48
I wouldn't call that industrial scale distillation, >> but sort of downream of >> there is some big gray area where it's like how do you actually define these? >> Yes.
32:57
And so defining that is going to be what that's going to be the conversation that plays out in in DC like behind the scenes on the basis of this and that's where the actual negotiation is going to happen between uh you know the the position of Nvidia and everyone that signed the letter versus the position of Enthropic and everyone who didn't sign the letter.
33:16
They're going to sort of decide okay well if you can prove this this and this and you can show us that your API was getting hit by these different things and you have a really solid report of what happened and then the model also you know sort of you know checks these boxes quantitatively when we do this eval then maybe we will pressure it and then what does that actually mean?
33:38
Um, you could go after the lab that committed the distillation attack with lawsuits, but that seems really difficult given the international nature of these attacks.
33:47
So, we're sort of back to where we started where, you know, you're you're like, what what can the government do that the lab can't?
33:55
Like, the lab should be looking at every customer and saying, "Oh, this seems like someone who's trying to distill.
34:01
They keep asking for basically what looks like a lot of training data.
34:03
They they're not acting like a normal user just being like, "Build me a website." Okay.
34:08
Um anyway, third he says all sufficiently capable models open and closed should go through mandatory mandatory safety testing.
34:16
So this was recently outlined by Dennis Hassabis over at Google Deepind as well.
34:20
And and it seems like that the two companies are in alignment on this in particular.
34:24
Uh and it's a somewhat reasonable position.
34:30
Uh although the risk is that small companies who have safe models could that aren't distilled could get tied up in a review queue for years before they can release.
34:38
can release. Like that would be very very annoying if you're uh recursive super intelligence for example and you don't have a Washington DC office and you're like hey we want to release our new model uh and they're like yeah totally like you got to go through the the the the review process get in line
34:56
and then it's like every you know every trillion dollar company is there with a ton of lobbyists being like review our model first because we want to get out a week before the the small startup and that's the frustration of biotech, the FDA, anything that goes through approval. We've talked about this with
35:10
We've talked about this with the nuclear stuff.
35:11
Uh it gets very tricky and so you want to avoid that and you don't want to wind up slowing down innovation that's happening on small scales and decreasing competition.
35:20
Uh Dario does do a good job of like acknowledging upfront that he says it would protect a US AI companies from competition, but that's never been my goal with anything that he's saying here.
35:32
Um, and so, uh, it's still worth working through what happens in a really adversarial situation.
35:38
Like what if a foreign lab distills a bunch of frontier models, they're the most aggressive, they're just distilling everything.
35:44
Then they jump forward a bunch in capability, uh, they get a bunch of smuggled chips, they take all the restrictions off of cyber, all the restrictions off of bio, uh, and then they just drop the weights on like a torrent or they put it up on Hugging Face, and Hugging Face is like, "This is really crazy. No one likes this.
35:59
There's a lot of pressure to take it down."
36:01
I don't know, but it's out there.
36:02
Uh, like what does the government actually do?
36:04
Like the government probably pressures or bans like hosting the weights, uh, maybe serving the model.
36:12
You maybe won't be able to run it in American data centers.
36:13
You go to the Neocloud and say like, hey, this thing is actually bad.
36:16
And I think people are divided on this because they see the current models not as actually dangerous, which is totally reasonable to assess that it's not that bad.
36:24
Um, but like if there was a model that was like, "Yeah, it's actually just like the killing machine."
36:30
Like I think most people would be like, "Yes, I'm democratically voting to not serve that because it's just like it's an annoyance at at best and like actually at worst."
36:41
>> And the other big question is like how much compute do you actually need for it to be dangerous, right? >> Yeah, totally.
36:46
>> Is is uh like having some GPUs in the back shed going to be enough maybe for sufficiently advanced model? Yes.
36:52
or do you need >> access to a ton of racks, >> ton of powert >> and as soon as you're a US-based company with a real data center with a bunch of NVL72s in there, uh you probably have registration and you know all sorts of just like business registrations where the government can reach out to you and say, "Hey, we're actually really worried about this."
37:18
Just like there are other things you can't host in a data center.
37:21
There's all sorts of stuff that's illegal.
37:23
It's just intellectual property. >> Yeah. Exactly. Yeah.
37:25
That's that's >> like you can't even just just because you have a data center doesn't mean that you can like take an open-source, you know, uh >> you can't you can't as a data center. >> Oh, yeah.
37:34
Open source Marvel like they'll be >> or even even even a you know a CRM company can't knowingly support like a organized cart global cartel that is like trafficking narcotics, right?
37:46
You have you'd have to imagine like uh that uh they have >> they have to vibe with their own >> balance it [laughter] >> maybe.
37:56
So uh so so uh what what's interesting is like what is the next step of that?
38:02
So if there is a bad model uh and and everyone agrees like okay yeah we got to not host this not distribute this like yeah the weights are out there people are trying to like sort of run it a little bit uh but does it go offshore?
38:14
Do we wind up in like the crypto scenario where there's like these offshore things and people are using VPNs to get access to it?
38:19
Like what level of aggression do you see from the US government in that scenario?
38:25
It probably should be proportionate to like the danger imposed by the model.
38:28
Like if it's just a model that's like that's like annoying or like slightly IP infringes, but like no one's really being like I'm not I'm I'm canceling my Disney subscription because this new model will generate me Disney IP.
38:39
Like that's probably not like okay, put up a crazy firewall.
38:44
But if it is like the ultimate hack machine that's like stealing everyone's money from the banks, then yeah, you are going to put up the the firewall and sort of be much more aggressive.
38:53
So I think the response will be in reaction to whatever the power of the models are.
38:56
But uh it'll be interesting to go back and forth.
38:59
Anyway, uh all in all, the letter clarifies a lot about the anthropic position.
39:03
So I think it's good that it it came out.
39:05
Uh but it's still worth working through the game theory of like what happens down the line.
39:08
Policy interventions is all we got here.
39:10
And I think it's still too generic at this point.
39:13
I want to know like what policy looks like. I want to predict that.
39:15
I want to understand uh what's actually being proposed, what people like, what people don't like.
39:20
Um and so I think we'll learn more about this in the coming days.
39:24
Let me tell you about Figma agents. Meet the canvas.
39:26
Your AI agents can now create and modify your Figma files with design system context.
39:30
We have Tay Kim in the waiting room.
39:32
Let's bring him in to the TVP Ultra Dome. Tay, how you doing?
39:38
>> Hey guys, doing great. >> What's going on?
39:40
So, so tell me, last time you were on the show, you bottom ticked it. What's going on?
39:46
>> Uh, I think I made the bullish call on CPUs, memory, and Nvidia.
39:48
Nvidia is up like 5 10%, but >> nice.
39:53
>> Uh, the CPU names have still doubled even after this big draw down, and the HPM names are up 100%.
39:58
So, I I'm hoping that, you know, it's the same thing again.
40:03
I come on here and stocks go up again.
40:06
Yeah, ideally we could have like an emergency reserve of take [laughter] appearances.
40:10
So if the market is ever downrate the strategic reserve we call you up, you jump on >> and then it's funny because it was literally the exact bottom and you know it went exponential after that >> take him effect.
40:22
So where are we right now with the level of FUD, the level of uh downward pressure on the AI trade broadly, the chips, the semi-rade?
40:31
Like reset for us on like where sentiment is and then we can work through the different pieces of counter examples.
40:40
So, I think sentiment's very negative.
40:42
We kind of had this >> huge up parabolic up move the last few months and likely a lot of retail and hedge funds piled in >> and we're seeing this unwind.
40:51
Now, I I think the first a big part of it was the Iran war getting worse.
40:55
Iran war getting worse. uh every time we had the >> the first ceasefire negotiations stock started taking off right right after that and then when you we had the actual ceasefire uh we had a follow through and then as soon as Trump started bombing Iran again you know chip stocks have kind of plummeted in the last two three weeks uh and then now we we're seeing just you know back to the old uh pattern
41:21
of uh media and the viral hot takes uh spreading a lot of FUD uh I think we saw earlier earlier this month um I think uh Reuters quoted like Zuckerberg's about aentic AI they took it out of context and then every media person was with a hot take that this met Meta was seeing bad returns and they're going to cut a capex and then we had leaks right after that saying that it looks like Meta is going to raise capex. So we're seeing a
41:47
So we're seeing a lot of this hot take FUD.
41:49
Uh yesterday I think we had a flurry of stuff that scared people.
41:54
um the Wall Street Journal vendor financing uh article that we'll see we'll see what happens with that.
42:00
We had the CMXT IPO in China and everyone freaked out over that.
42:04
We had the information article on ASML.
42:06
Uh we could go through each one and then the Kimmy thing.
42:10
It's obviously a big thing.
42:12
>> Yeah, we'll definitely get there and I want to talk about open source and Nvidia's strategy there obviously.
42:15
Uh starting with the Mark Zuckerberg news in Reuters. This was July 2nd.
42:19
Meta's Zuckerberg says AI agent tech progressing slower than expected.
42:24
Um Zuckerberg added that the company's reorganization that included major job cuts was not as clean as it could have been.
42:32
Zuckerberg and other meta executives have been seeking uh to moderate some of the c of the organizational changes introduced this year and they said that uh the trajectory of agentic development over the last four months hasn't really accelerated in the way we expected.
42:47
accelerated in the way we expected. uh the company's bets on new structure haven't come to fruition yet and so people were sort of reading this as maybe Meta is going to pull back but then it felt like the response was extremely quick with with Bos going on a
43:04
podcast and Alex Wang sharing a whole bunch of progress across um a few different uh models and data points and then semi analysis wrote a whole bullcase for MSL talking about how they have compute and also So they have uh
43:20
more of like the internal structural alignment to sort of properly yolo in the AI era if I'm boiling it down as brutally as possible just because with Google there's always this debate between oh do you sell the TPUs or do you sell the cloud comp do you have
43:37
bended in the product whereas Mark Zuckerberg's able to sort of like go all in on this new idea and so maybe there's more more glimmers of hope there but uh what else have you been tracking downstream of meta those ambitions. >> Well, I mean, they've been very upfront
43:51
>> Well, I mean, they've been very upfront that they're investing heavily in AI.
43:54
Alexander Wang is tweeting multiple times every few weeks that they're they're going full force.
43:58
They're going to redo open source AI models.
44:00
They I think he said that at the YC event over the weekend. >> Yeah.
44:06
>> And it's I mean, if you actually look at and then Reuters came out, I think with an article saying that they're actually going to raise capex dramatically uh this year and next year.
44:14
So all that kind of fear that that quote about adventic AI from the town hall um that kind of like spooked the market for a few days, it kind of it was completely false.
44:26
Um, >> it feels like it's a Yeah, it feels like it's a comm's air because the the language that's been coming out of meta has been a little bit like AI is going to replace our employees and it feels like it'd be much better for them to to come to the market with a message of we're going on the offensive like we're a hyperscaler.
44:47
>> Well, to be to be fair, that that was an internal town hall.
44:49
They didn't mean to leak it and Reuters leaked that one quote and put out put out the headline before the article.
44:55
Yeah, it's it's interesting like Meta did Meta basically go through like an 8-year period where like internal town halls didn't instantly leak.
45:02
>> I think everything leaked always.
45:02
I think I think everything's been >> I know.
45:05
But there was there was a period where like the the the the sort of attention of the media was way way way less on like what Meta was doing internally relative to the 2010s >> and all of that attention just went to the labs, right? >> Yeah. Yeah. Yeah. No, that makes sense.
45:22
Um yeah, I I guess the question is like the question that I keep coming back to is like where is their revenue ramp?
45:32
Where is their AI revenue going to ramp and when? Right?
45:34
Because as you >> super would say ads like the ads like the AI has they're accelerating there.
45:43
>> That's always been my view too.
45:43
But when you're when you're continuing to ramp capex >> Yeah.
45:48
with and saying like we're going all in on aentic and we're building a harness and we're also going to do open source and it's like well what is the strategy? >> Sure.
45:58
>> Like what's the breakthrough?
45:58
What is going to take you to a billion dollars of like pure AI product revenue or or or just API revenue?
46:06
just API revenue? what's gonna take and then to five and 10 >> and what's gonna allow you to like justify the spend other than I think the market would love if they just said yeah we actually need all these GPUs because we can actually be 10 time we're already good at ads we could be 10 times better and that's where we're going to get the
46:24
the ROI on all this capex >> well well they're definitely getting ROI on that the market is worried about you know all this extra capex on the they're going for the frontier AI model race again >> and they had to reset what they Yeah, >> they had to a lot of people left and now Wang hired a ton of people and uh we'll see what happens over the next it's going to take time. It's going to take
46:44
It's going to take six to 12 months before we see any more progress.
46:47
But that model that came out a few weeks ago um was a lot better than people expected.
46:52
It wasn't >> you know the frontier but it was much better than what people expected. >> Yeah. Yeah.
46:57
Uh so uh how have you been processing the NVIDIA letter around uh open source and all the back and forth all the people jumping on the companies that have been staying back how you work through that?
47:14
>> It's been very impressive what they've been they basically united the entire tech industry against Enthropic in in the last like three four days.
47:21
18 trillion in market cap has signed on.
47:25
Last time I checked across uh >> Google Amazon signed on eventually the only >> Yeah, they signed on yesterday.
47:31
They tweeted out interesting.
47:34
>> I think Apple is still the hold out which is kind of strange because they're the one that would most benefit from open source open weight models being you know more available I would think.
47:44
But I would I don't know what Apple, >> but I mean they pretty much got the whole tech industry to uh kind of corner anthropic in their position. Yeah. >> Um Open AI signed on. >> Yeah. What did you think of?
47:59
>> Obviously, Nvidia is afraid. >> Yeah. I don't know.
48:01
I don't know how how if I don't know if they're really >> I don't read I don't read it as being like cornered by any means. >> Right.
48:12
Well, Jensen is on the record that, you know, he said, I think to Bloomberg, that there was rising sentiment that something was going to happen on the on the regulation front at White House or whatever.
48:23
So, this is we saw that this was this was last week.
48:28
You had at least four people in the admin >> say, "We're not against open weights.
48:33
We're against distillation."
48:33
And at least I was reading into that of some type of regulatory action around open weights and and then positioning it as we're targeting this is this is like wrote yesterday. Yeah. Yeah.
48:47
>> Um about you know push back and restrictions and he's done doing it under the safety umbrella but u definitely Microsoft and video are worried that uh the white house or congress is going to do something on this front and that's what
49:03
>> they uh it seems very reasonable that he would know have no problem with like Gemma or Llama or any of the open source from like American hyperscalers where if you find out that they're distilling you just walk across the street and sue them. Uh, and also these big companies
49:14
them. Uh, and also these big companies have huge huge I mean they have safety teams but also just like huge incentives to not have a safety incident happen on their watch because you're trying to like catch up to the frontier and then all of a sudden you have a safety incident that's going to be really bad
49:30
for your overall brand uh and you have a different business to protect whether it's social networking or uh Google search if all of a sudden the Gemma model winds up being a thorn in someone's side for a cyber security reason or a bio and uh that would be uh really really bad. But uh a foreign
49:46
But uh a foreign company that is just like hurling it over here can kind of just be like you guys deal with the consequences potentially.
49:54
So I think that's what what what Dario is worried about.
49:55
Uh what about um the overall idea of like we're we're it feels like we're sort of replaying the deepseek moment.
50:03
Open source is going to reduce cost and so that's a reason to pull back on the AI trade overall.
50:11
How have you processed that?
50:12
It's almost it's almost a perfect catalog.
50:14
Uh people are worried about Kimmy. Yeah.
50:16
But when you actually read the technical paper and their blog posts, uh this is not a tiny efficient model. This is 2. 8 trillion parameters.
50:25
It's going to require a ton of compute to serve.
50:27
>> I mean, we saw the first day they put it out that their servers got slammed.
50:31
>> Um even in the blog post, they say it's best run on a a kind of a a server with 64 GPUs.
50:38
So big super clusters that areworked well and and that's perfectly runs great on Nvidia.
50:43
And if you remember, you know, during the whole Deep Seek thing about a year a year or so ago, uh the market freaked out that, you know, Deepseek was going to was so efficient that it will lead to a compute glut.
50:54
But Deepseek was the example of the reasoning model that actually it was the opposite.
50:59
It created a ton of demand.
51:00
And I think the same thing is going to happen with Kimmy where uh when you have more capable models that come out, people find uses for them.
51:07
And right now, just like last year when reasoning models took off, Agentic AI and agents are taking off right now.
51:14
And the market is kind of like not realizing that because right now, just like last year when reasoning models were taking off, right now Agentic AI is taking off and the next six nine months are going to be bigger than anyone believes.
51:27
And Sam is on the record.
51:29
Sam is on the record over the weekend saying >> at the YC thing again.
51:36
Like people don't I don't know why people don't like listen to the it's on YouTube. Yeah.
51:42
>> That the next six months it's going to be much more dramatically better for AI than the last two years in terms of advances capabilities.
51:48
And I I heard you say RSI before.
51:50
I think it's going to be RSI.
51:52
Uh people inside Open AI and definitely anthropic.
51:55
definitely anthropic. anthropic blog post on this RSI I think is a lot closer than people think and if RSI actually happens in the next 369 months that's going to soak up insane amount of I mean we have this exponential ramp for
52:10
reasoning exponential ramp for aentic and then if RSI actually happens and I think it it sounds like both frontier labs think it's going to happen very soon that's going to soak up an unbelievable amount of compute as the AI
52:22
models you know use more compute to to self-develop and improve and I think that's one thing missing >> that both anthropic and openi are kind of winking that oh it's happening anytime I tweet something on RSI all these frontier AI researchers [laughter]
52:38
like my tweet so I think that's good >> uh what what is your what is your sort of framework around compute hoarding because certainly certainly it is it has been happening when you look at when you look at you know uh like going back to the meta example Right? They're not
52:55
They're not selling compute yet.
52:57
They're maybe curious about it or or exploring some deals, but they have all this computing on their own ability to create the capability that will have enough demand to uh justify justify that.
53:09
Do you just think it there there's so much demand overall that it just, you know, even if there's there's hoarding, it just will leak out >> and it's okay.
53:21
>> Well, there's so much demand overall.
53:21
I mean, the SKH Highex executives said uh during their IPO run um that their customers are asking five to six times more than they're able to serve and they're going to double capacity over the next five years.
53:34
They said and their customers and I'm going to assume it sounded like Jensen um are asking for five to six times more than they're they're able to build.
53:44
So, there's overwhelming demand.
53:45
You guys were at the advanced AI uh AMD event.
53:47
Lisa Sue raised her CPU, Agentic CPU forecast.
53:55
Just three months ago, it was 120 billion for 2030.
53:57
Three months later, they raised it to 220 billion. >> Yeah.
54:02
>> Like, she doesn't do that. She doesn't do that.
54:05
>> You have that just on the >> I've got that ready.
54:07
I can do [laughter] whatever.
54:08
>> I mean, like, >> well, I just love this chart because he called it perfectly. >> He actually did. It's crazy.
54:16
CEOs don't don't you know raise their TAMs by like these multiples in a few months if they're not seeing insane demand coming in.
54:25
>> Especially not public CEOs who are serious business leaders who've been running like non-me stocks for decades and are serious.
54:33
>> Everyone's freaking out that this is like the dotcom bubble all over again. Don't do finance.
54:38
>> But what if these hyperscaler GPU cloud businesses are amazing businesses?
54:41
Like Morgan Stanley says, if you do inference, >> it's 60 to 80% profit margins, right?
54:49
These are amazingly profitable businesses as long as we keep growing the next few years.
54:53
And and again, [clears throat] just like last year, we're we're on this exponential run right now over the next two quarters.
55:00
And the market isn't seeing that.
55:02
Everyone's freaking out that oh no, we're spending too much.
55:03
And even uh Sam Alman uh podcast came out today and another podcast. Sam is out there.
55:12
He said that he regretted, you know, pulling back on the compute purchases.
55:16
They made a mistake by uh not putting the pedal to the metal because now things are taking off again.
55:21
So >> like I I like Amazon, the CEO in April, if you everyone read his annual letter Andy Jasse wrote, >> he talks about how free cash flow works.
55:34
We're not betting $200 billion on a hunch. We see the demand.
55:36
We know it's going to be insanely profitable.
55:39
and free cash flow positive in the medium to long term.
55:44
So that's why you're investing $200 billion now and in a year or two we're going to see insane amounts of free cash flow.
55:50
The thing that people are worried about right now, it takes time to build out these data centers and fabs.
55:55
And you that now hold like if you if you see uh if you see free cash flow that that that assumes that uh like like the revenues have to catch up and then the capex can't grow more exponentially.
56:10
And so that means you have to see some sort of plateauing.
56:14
Maybe it's at the end of the chart, maybe it's this 2030 range, but there is a different world of just like continued growth forever and then we sort of like run out of money.
56:23
>> The push back I have there is that's a static view, right?
56:25
If they don't grow revenue >> for the next three years, yes, you can't do that.
56:31
But they're growing Azure is growing 40%.
56:34
Google Cloud is growing 80%. >> Yeah.
56:37
>> You know, Amazon's growing >> high double digits.
56:40
So if if if revenue is growing 40 to 80% this year, next year, and the year after, that's more revenue you have, that's more operating capital you have to invest, >> right? >> Yeah.
56:55
>> So that's that's what people are missing.
56:56
missing. And if this stuff if if the data center you're building now you're spending all this now generates unbelievable free cash flow in 12 to 18 months because you know this agentic AI is actually changing and rearchitecting all the workflows inside companies and
57:14
you need to do the agentic AI coding agents to make your product better because if you don't if you don't iterate a 100 different iterations of your product in R&D if you don't do AI just like AT&T is doing at the gent uh advancing AI at AMD. He's talked about
57:28
He's talked about they're putting a 100 Gen AI models into production.
57:33
They're burning a trillion tokens a month and then that's growing double digit.
57:37
The reason why they're doing that is because you by using a AI you're providing a better customer service, a better you have a better product R&D and you're helping your companies make better products and services.
57:49
And if you don't incorporate AI into your company, Verizon, your other company is going to do is going to incorporate AI and then disrupt you and then you lose all your revenue. Yeah.
58:01
>> So, everyone's worried about ROI.
58:01
ROI is important, but you also need return on revenue because uh if you don't use AI, your your rival is going to use AI to beat you in the market. >> Yeah. Yeah.
58:12
You know, I I I think the diffusion story is still even though we got like sort of jitters by the token maxing thing, just the actual usage of AI across companies is still pretty limited in terms of the amount of people that are using it, the time that those people are using it.
58:27
Like there definitely is a San Francisco bubble of startups where everyone is using AI a lot.
58:33
But if you just walk into a normal business, a lot of people are like, "Yeah, I got to check that out."
58:38
Which is a >> Let me give you some some context here.
58:42
Yeah, Ara Karazzian uh Jared Sleeper over on ECU saying uh enterprise adoption disparity remains enormous and he cited Ara saying >> uh usage would 100x if every company adopted AI to the degree of the most advanced companies.
58:57
advanced companies. There's like this there's a small group of companies that are >> people forget in the ramp in the ramp data like adopting AI can mean like having a chatbt pro account for someone which is like not exactly the same as like using codecs and like coding agents
59:12
and stuff like it's important I think that you know if I have someone on my team I want them to be able to go and do a deep research report but uh that's like table stakes the question is like are you actually speeding up anything that's repetitive in your job and uh that diffusion is just starting to take hold. >> So, so the total market size in terms of
59:29
>> So, so the total market size in terms of IT and knowledge management in corporations, it's about $6 trillion, right, a year.
59:39
>> Uh the two main frontier AI model companies, Open AI and anthropic, I'm going to say I think this is roughly accurate are doing $120 billion combined in ARR. >> Yeah.
59:49
>> You know, why can't that go to 200, 300, 400 billion in the next year or two?
59:52
I mean, they're growing at exponential rates. Yeah.
59:56
when we're taking off and the if the market is $6 trillion, right?
1:00:01
Why can't they grow to 200, 300, 400 billion in the next couple years?
1:00:03
I mean, it's like just do a little logic and rational deduction. >> Yeah.
1:00:10
>> Um this is definitely possible and it's happening right now and it's accelerating and people aren't, you know, they just taking, you know, these big headlines where we had this uh you know, $50 billion from the Financial Times and we find out it's over 30 years.
1:00:25
It's like on the homepage about >> wait. Yeah. Yeah. Okay.
1:00:27
I wanted to ask you about this.
1:00:28
Nvidia revealed this tenant for $50 billion data center that will use its chips.
1:00:32
Explain what is actually going on here.
1:00:36
>> So, the Financial Times, you know, put on their homepage today that Nvidia is going to backs stop a lease for a data center in Texas for $50 billion.
1:00:43
And I saw that I was like, "Oh my gosh."
1:00:45
Oh, that doesn't sound good.
1:00:50
>> No, it literally sounds like they're buying their own chips.
1:00:51
So, it sounds like the most bad thing you could do. >> Yeah.
1:00:56
[laughter] Then they actually read the article, like halfway down the article, it's like a 15-year lease.
1:00:59
Uh, and it's only $50 billion if they renew the lease after 15 years.
1:01:05
So, it's like over 30 years if they renew it.
1:01:08
Then, if you think about that, you're like, "Wait a minute, 50 billion divide by 30 if they renew it." That's like Yeah.
1:01:16
Nvidia's 15year lease commitment for the Texas site is worth basically 20 billion.
1:01:20
and renewal options would take the total value to 50 billion over 30 years according to HUT 8. >> Okay.
1:01:28
But what do you think their what what are their plans for the site?
1:01:30
Is this >> they are going to have some like what do you expect them >> so so my my point is this is a billion you know whatever a billion or$2 billion dollars a year right it's a non-story but it's a it's a big headline sensational headline on the homepage. >> Yeah. Yeah.
1:01:48
And also and also it's not that it's it's not like you're taking a $2 billion loss every year.
1:01:53
It's you are the tenant and then you are also renting that out.
1:01:57
So hopefully you're making profit.
1:02:00
>> It it's a rounding error.
1:02:00
It's like, you know, they're doing 320 billion run rate a year.
1:02:05
Now that's going to go to 400 500 billion next year.
1:02:08
And we're talking about something that might be a billion.
1:02:11
you know, like this is not a story, but you this is how people run with the sensationalized headlines and people panic and freak out.
1:02:21
>> I think they just wanted to say the biggest number.
1:02:23
>> Yeah, >> that's exactly the point.
1:02:24
And and we're going to see what happens with this Wall Street Journal article.
1:02:27
Um both Open AI and Nvidia are not commenting so far. >> Uh we'll see.
1:02:33
>> But take us through the rumor. >> Have to wait. >> Rumor will tell.
1:02:35
>> Well, it's not a rumor.
1:02:35
It's the Wall Street Journal and other uh people reporting. Yeah.
1:02:38
Nvidia is in talks with open AAI to backs stop soft bank up to 250 billion you know billion we don't know the details and I don't want to speculate and comment but let's actually see the details before we I I think the market had a really big negative reaction yesterday to the story uh because everyone I mean Jim Kramer was telling his audience like sell everything at the open today because AI and data fenders.
1:03:05
com you know it was insane it's just let Let's see the actual deal and the metrics and the numbers before we panic and freak out. >> Yeah. Yeah. Yeah. Yeah. That makes sense.
1:03:19
>> So, honestly, when you say freak out and sell everything, sell your dollars, sell your h sell your house, sell your stocks, then I'll freak out.
1:03:29
>> But until then, Tay, I feel I feel okay.
1:03:32
>> I I mean, I just see the fundamentals.
1:03:32
I see the CEO of AMD expanding her TAM, you know, dramatically in over the last 3 months.
1:03:39
I see RSI under horizon like every AI researcher is like, "Oh my god, this is going to happen.
1:03:45
We have to get there sooner."
1:03:46
And then I see, you know, the obvious use case of uh Aentic AI where you have to rearchitect your workflows internally.
1:03:55
Every company has to do this.
1:03:57
So everything is taking off.
1:04:00
You see like um you see uh when the president of Korea came to San Francisco area last week, you know, they had like a day in the valley instantly.
1:04:08
Nvidia CEO Jensen Hang Broadcom CEO Hawkan Daario, you know, Sam Alman are there, right?
1:04:20
You know, do a little logic deduction.
1:04:22
Wh why are they there like crazy?
1:04:24
Because they need HPM memory and they're dying to have it.
1:04:26
So if you think about that, that means there's insane demand and HBM memory is in in >> in shortage.
1:04:34
There's tremendous demand for it, right?
1:04:37
>> Talk about the Oh, sorry.
1:04:37
Talk about the Nvidia CUDA mode.
1:04:40
It feels like uh a big piece of AMD's advanced AI event was uh maybe the CUDA mode isn't as much of an issue anymore in the age of agentic AI.
1:04:54
you can have an AI agent write you the software that you need to use any chip and that creates uh less pricing power for Nvidia, but there's another world where you're not really like Nvidia doesn't necessarily need a moat because everything's just growing so fast that they're still growing.
1:05:10
But how have you interpreted the processing of like the potential death of the CUDA moat? >> So AMD is on it.
1:05:17
Kimmy wrote like a couple paragraphs in their blog post about how they created a GPU kernel, all that.
1:05:24
So, everyone, you know, sure, >> get scared or whatever.
1:05:28
>> It it's like we'll see what it's like in real life.
1:05:31
You know, this is just, you know, AMD is incentivized to say, "Oh, Kudo is not a problem anymore."
1:05:37
>> CUDA's been a tremendous moat and I think it continues to be a moat.
1:05:39
And the reason why is uh it's super reliable.
1:05:44
all the bugs have been optimized and fixed and that comes from hitting the software you know millions of times and billions of times right like you don't know if you know if you use cloud code or Kimmy that what they figure out using their training data is going to work in
1:06:01
the real world right they could talk about one little piece that does well let let's see how it actually works yeah but Nvidia's big mode is uh its scale it's co-design of actually working through the networking, the CPU, the GPU, and how everything works together. And the other big thing is their balance
1:06:18
And the other big thing is their balance sheet and their ability to uh get supply commitments from, you know, I like I think I said this before um optical startups are like upset because Nvidia secured all the supply for all the optical components.
1:06:34
Same thing with TSMC wafer.
1:06:36
Same thing with HBM memory.
1:06:36
So the Nvidia is using their size and gorilla and be able to prepay and get components that are in shortage.
1:06:44
So they've become the dominant, you know, over the next year or two, you're going to see Nvidia able to add tons of revenue because they were able to lock up all the supply components.
1:06:53
That's another thing that people don't really talk about is their supply chain and their ability to work with partners and secure secure component inventory.
1:07:00
Is there still energy FUD that we would run into an energy bottleneck before we run into a chip bottleneck?
1:07:07
So, >> so Jensen said this last week on the Bloomberg interview that there are a lot of bottlenecks including um data center shell power and all those things uh components um energy whatever.
1:07:19
So all those things it sounds really bad, right?
1:07:23
And then right after that he said I think we have the chip industry has enough supply to double their revenue every year.
1:07:30
basically implying Nvidia has enough supply uh for for energy and all that stuff.
1:07:35
No one is pricing that in.
1:07:37
So, everyone talks about bottlenecks.
1:07:39
Nvidia CEO just basically told you on Friday that they have enough uh enough supply chain and all the bottleneck stuff to double revenue every year.
1:07:49
And no one, you know, Nvidia's revenue uh estimates for next year are a lot lower than double. I'll tell you that.
1:07:57
Do you think the market uh prices in just how much of almost every important AI company in every category Nvidia actually owns?
1:08:07
actually owns? like it it it feels like every single like we're constantly focused on who's going to raise capex next and and and and where's this quarter coming in and it feels like in two or three years people will look at Nvidia's balance sheet and be like wait they have what what I imagine then will be you know could could end up being a
1:08:28
trillion dollar plus of just like ownership and all of these great companies which again just goes back to the advantages of that early scale while they're, you know, while while these companies are trying to compete away Nvidia's margins and all these different things, they've been able to accumulate again positions in in all of these incredible companies. I mean, we saw the
1:08:48
I mean, we saw the SSI news uh yesterday is a great example of that.
1:08:54
Um, but but how do you look how do you see it?
1:08:58
>> So, I I think look at Jenison's history in investing in these companies and core and see how much money they made.
1:09:03
uh they just bought stake in uh the optical companies momentum and co coherent.
1:09:07
Um Jensen is enabling the future because he sees this overwhelming title of demand and he needs these companies to be able to build up their supply chain uh and to give uh supplies and chips to to Nvidia.
1:09:24
So they actually ramp very hard.
1:09:24
Um you know everyone's freaking out that this is vendor financing.
1:09:28
What if hypers scale GPU cloud is so profitable and these companies need capital to build up that supply so they can uh serve the GPU cloud uh services over the next year or two.
1:09:42
Maybe Jensen sees that coming like he did with all these other companies like Cororeweave and uh that's why he's investing in these companies to be able to expand their ability to make the components uh the industry needs.
1:09:54
So I I think you're exactly right.
1:09:56
In in a year, two, three years, uh Nvidia is going to have like all these stakes in these companies and it's going to look like uh he was a good investor.
1:10:04
Um cuz he has been in the past.
1:10:10
>> I mean, I mean, the man can buy a leather jacket for like five grand and sell it for a million dollars.
1:10:13
I don't know what else you need to see.
1:10:16
>> I mean, think think about >> Are you the secret bidder? Did you win that? >> No.
1:10:20
>> You got to get you a jacket.
1:10:20
Uh the real the real question is I do >> how long until Someone distills a jacket and open sources it, you can get a dupe of a Jensen jacket for two bucks. That's what I want.
1:10:33
Anyway, thank you so much for coming on the show.
1:10:34
Jordy, you got anything else? >> Uh, this was great. >> Yeah, this was great.
1:10:39
>> Thanks for Thanks for putting up with all of our jokes.
1:10:42
>> Yeah, >> hopefully this becomes the lucky charm for the markets. >> Yes, I agree. [laughter] >> I agree. I agree. >> We'll talk. >> Bottom is in. Great to see you, Tay.
1:10:52
>> Have a good rest of your week. Goodbye.
1:10:54
Let me tell you about CrowdStrike. Your business is AI.
1:10:55
Their business is securing it.
1:10:57
Crowd Strike secures AI and stops breaches.
1:10:59
Um, I wanted to run through on the power issue.
1:11:02
There's an interesting article in the journal.
1:11:04
An underground nuclear reactor is coming to this Kansas town and it's dividing locals.
1:11:09
This is something that I had tweeted about years ago, like why don't we just put the nuclear reactors underground?
1:11:15
Uh, put solar panels on top.
1:11:16
Best of both worlds, optimal use of energy.
1:11:19
Uh but there's a lot of fear and uncertainty and doubt about this one.
1:11:24
They say no one has tried operating commercial one a mile down until now.
1:11:25
Uh it's great that it's here.
1:11:29
It's kind of bad that we're the guinea pigs, says the residents of Parsons, Kansas.
1:11:32
Residents of this sleepy farming outpost agree on many things, but whether to put an experimental nuclear reactor a mile deep in the granite beneath their town isn't one of them.
1:11:45
elected officials and some others see a chance to create jobs and lure data centers and manufacturers to a rural patch whose economy has been flatter than the surrounding corn fields.
1:11:55
Another group is effectively saying not under my backyard, it's ni, not nimi because it's not under my backyard, ni or something like that.
1:12:04
Uh, I put $125,000 into my house and now a nuclear reactor is coming to town, said Gerald Johnson, an IT professional who plan to retire in Parsons.
1:12:14
I can't think of a worse idea.
1:12:16
No one has tried operating a commercial nuclear reactor deep underground until now.
1:12:20
I'm surprised no even like the Soviets in like 1950 didn't try it.
1:12:24
I feel like they were trying everything. I'm surprised. Yeah.
1:12:28
Uh, but the so-called gravity reactor is is the creation of Liz Mueller and her father, Richard Mueller, Emertus professor of physics at the University of California, Berkeley.
1:12:38
It's Berkeley people again. And an inventor.
1:12:40
They founded Deep Vision, a three-year-old California startup that raised $150 million in the past year, including $40 million last month through through an IPO uh largely to fund the work in Parsons.
1:12:51
Parsons with a population of 9,600 people sits about midway between Kansas City and Tulsa, Oklahoma.
1:12:58
Deep Vision drilled a first test hole this spring on a 100 >> office chairs like that have really fallen off.
1:13:03
Which tells you >> now might be the time to bring them back. No, >> right.
1:13:09
>> I need I need a new office chair.
1:13:09
I might go for one of those. The highback leather. Uh it's good. Good.
1:13:12
uh deep vision drill the first test hole this spring on 100 acres at a mostly overgrown industrial park dotted with old munitions bunkers just outside town.
1:13:23
On a recent day, Maurice Leaf Fountain Deep Vision's senior engineering director showed off a pink fleck granite retrieved from the company's first test hole and joked that the billion-year-old rock would make a nice countertop.
1:13:35
An empty steel container canister sat on a clear drilled pad uh waiting to go down a second hole.
1:13:43
This year, the plan is to send another one loaded with nuclear fuel into a third hole to heat water a mile underground and generate electricity on the surface in 2027 2028, an an astonishingly short time frame by industry standards. Uh, interesting.
1:13:59
Anytime you're putting a nuclear reactor in the hole, it's kind of scary, he said in his office.
1:14:02
It's great that it's here, but it's kind of bad that we're the guinea pigs. Uh, very interesting.
1:14:07
I'm surprised we haven't heard more about this company, this idea, uh everything that's actually being planned.
1:14:13
There's something a little I understand where they're coming from.
1:14:15
There's something a little bit nerve-wracking about uh like even though you would think a mile deep if something goes wrong, it's less of an issue, it feels like well people it's harder to get to and like just go and solve the problem, deal with it as opposed to like oh yeah, it's a building over there.
1:14:30
I see people coming in and out all the time.
1:14:33
The experts are in control. I don't know. Um, what do you think?
1:14:35
Are you pro nuclear underground?
1:14:37
A mile underground could be the future.
1:14:42
>> Could they not find maybe a place to do that that wasn't right under a town?
1:14:48
>> It's not right out of town. It's outside of a town.
1:14:50
You need some infrastructure, you know, like it and I I I bet what they would say is like, look, it's it's a 10,000 person town. We went miles away.
1:14:58
We're on a 100 acres of land.
1:14:58
Like we are outside of the town.
1:15:01
But yeah, there there aren't that many places that are truly like uninhabited for like hundreds and hundreds of miles just because of the nature of of America.
1:15:12
There's there's towns all over the place uh every street and you need roads to be able to deliver equipment and whatnot.
1:15:17
Anyway, let me tell you about Codeex.
1:15:19
Codex is a powerful workspace for getting work done with AI agents.
1:15:21
Whether you're writing [laughter] code, analyzing data, creating content, or automating business workflows, Codeex helps you move projects forward from start to finish. What is this? How did we get here?
1:15:33
Anyway, we have Ben Zig from Revelio Labs coming on the show. How you doing, Ben? >> Good. Good. I love that intro.
1:15:42
>> That was just for you.
1:15:42
We're testing that out for the first time. >> Yeah.
1:15:45
Uh >> nice and matches the vibe. >> Yeah.
1:15:47
The vibe of the labor market.
1:15:47
Take us through uh a little bit on your background, how you work, and then uh some of uh what you're tracking in the labor market and how it ties to your actual business. >> Yeah, for sure.
1:15:59
So, so I'm a labor economist, been tracking labor market data for a long time and started Rebellio Lab.
1:16:05
So, Revelia Lab is a workforce data company.
1:16:06
We're collecting, curating, synthesizing all labor market related data. >> Yeah.
1:16:12
>> Uh that's out there in the world.
1:16:12
And of course, you know, a big question is how is AI affecting the labor market? >> Of course.
1:16:17
>> Though it, you know, we're we're uniquely positioned to answer that question and it's on everyone's mind. Mhm.
1:16:22
>> So we we started putting out this labor market this kind of AI labor market tracker which is really about answering like how is AI affecting the labor market today.
1:16:31
So not really getting into the speculation of what might happen. >> Yeah. Yeah.
1:16:35
Just >> but really like what do we know >> really quickly uh what is your business model?
1:16:39
Who who gets value out of this data?
1:16:42
And then I also would love to know uh how do you go about getting more accurate data because I see like you know obviously the the census bureau the government has access to do polling ADP is a very logical place to get data because they run payroll so they can see the data but uh what's been your strategy there and then who's the customer? >> Yeah. Yeah.
1:17:02
So I'll start with the customer.
1:17:03
So so a lot of it is hedge funds.
1:17:06
So they they're speculating on the performance of companies. [applause] >> Yeah. Nice.
1:17:09
Thank you for your >> cheer for hedge funds.
1:17:11
Um yeah, they don't get a lot of love these days but um but yeah they they are speculating on the performance of a company that they have no affiliation to.
1:17:18
So you have to understand the you know what's happening in the company the workforce dynamics >> uh HR departments for benchmarking also.
1:17:26
So strategic workforce planning people analytics talent intelligence these are all like kind of segments of analytical HR >> and academic research.
1:17:35
>> So it you know they of course want to know what's going on.
1:17:37
So, so these are so basically we we get the data not through surveys, not through payroll, but really from the internet.
1:17:43
So, you know, LinkedIn profiles, job postings, glass reviews, layoff notices, immigration filings, freelance platforms, like >> anything and everything that is in the public domain and that information has to be, you know, enriched and synthesized in a smart way.
1:17:58
like there's all sorts of sampling biases, there's lags in reporting, there's like raw text, you know, so we have to classify that to occupations to skills, >> seniority levels and you know, importantly work activities which is more of a recent thing for us but >> important these days.
1:18:14
>> And then I sort of back test against like the historical actuals to see that the models working and then you can be more up to date.
1:18:22
>> So we don't back test against financials because I trust it anyway.
1:18:25
I mean about like like if you ran your model on like what was the employment rate in 2021, you could look at the actual employment rate to sort of calibrate that your system is predicting employments correctly.
1:18:38
Is that is that roughly correct? >> Uh yes and no.
1:18:40
I mean for some models we can see what was retroactively revealed.
1:18:43
So you know when someone changes their job they don't necessarily update that right away and we can see that you know every time stamp has like more information than it did before.
1:18:51
>> So that's like a solvable problem.
1:18:51
But in terms of you know saying what's happening in the labor market at large >> we can kind of use BLS data.
1:18:58
So BLS is a Bureau of Labor Statistics.
1:19:01
We can use that data to kind of like proxy for it.
1:19:05
But that's got issues itself.
1:19:05
So I don't know if we want to use that as ground truth.
1:19:09
So you know I think BLS has you know view on what's going on from survey data ADP has from payroll data and we have from internet data and they're all kind of independent in their own way >> kind of uncorrelated errors. >> Okay.
1:19:21
So I want to understand how how you look at your data in the context of AI diffusion right so a company an individual company or an industry might have uh fluctuating like labor data right maybe they're adding a lot of people uh but then individually if you look at those companies maybe some companies are adopting AI quickly some comp companies in that sector aren't really adopting AI at all or they're doing in a very minimal way.
1:19:50
Let's say they just have a basic, you know, chatbt $20 a month subscription.
1:19:54
So, like how are you uh I was I was talking to John maybe was it 6 months ago?
1:20:00
I was saying like I really want there to be a firm that is just studying AI diffusion in specific industries and getting into the nitty-gritty probably doing surveys to actually understand how because every company says they're adopting AI but we all know that there's like such a broad spectrum and then of course some people are saying that just because they want to be uh in in feel like they're a part of the club. >> Yeah. Yeah.
1:20:29
Yeah, I think it's probably mostly those that want to be part of the club, but I agree.
1:20:32
I mean, so, so there's a few ways to get at adoption data.
1:20:35
So, so I think adoption is the hardest part of all of this >> because that's really a firm level, you know, piece of information, whereas AI exposure is like more of a person level piece of information. >> Sure.
1:20:46
>> So, >> so I'll tell you the way we do it in a couple ways.
1:20:49
So, one is that we we had a partnership with uh we still have a partnership with RAMP.
1:20:53
So, I know friend of the pod. >> Let's go.
1:20:56
Um so they they can track adoption um just using like AI spent. >> Yeah.
1:21:03
>> So they can see like dollars spent on tokens etc.
1:21:06
>> So so that's like a pretty good way to get adoption.
1:21:08
The problem with that is that >> first of all it's like a self- selected sample you know ramp skews toward more like tech.
1:21:16
Um which is fine like that's overcomeable.
1:21:19
The other issue is that it's anonymized.
1:21:21
So they can't release information at the firm level.
1:21:23
So you know when we collaborate with them like we have you know the labor market data and they have the adoption data.
1:21:28
So you know it's like complicated you know we have to send the data they have to like run something we have to do some matching.
1:21:36
So it's like a little bit it's got some friction.
1:21:39
>> The other way to do it is through um we use this measure which is used in a paper um re a recent paper um that measures adoption by like um sort of hiring AI integration teams.
1:21:50
So the thought is that you know if someone's like hiring AI integrators you know beyond some threshold that they're like taking it seriously >> and they're embedding it into their business processes.
1:22:02
And by that metric we see about 9% of firms >> like getting very serious about AI.
1:22:07
It's a very conservative way to measure AI adoption. >> Yeah.
1:22:11
>> But seems to be pretty good.
1:22:11
Like it's correlated with all sorts of other things. >> Yeah.
1:22:16
I sort of hate that idea as a metric, but it probably makes so much sense in larger organizations that that is a great signal, but it just feels like completely the wrong way to go about actually changing a business.
1:22:27
Like uh I I feel like adoption should be so much more ground up than like, oh, we're hiring a special team to do this.
1:22:34
But that's the way businesses work.
1:22:36
And I think you're correct to identify that and probably is very indicative of a change in the stance of the business.
1:22:42
Where's where's an area that that AI is really good and you're seeing uh job loss? Hm.
1:22:51
because like AI is pretty good at software engineering now or generating code and and you know the companies that are adopting it the most are hiring a lot of engineers whereas I've heard in LA specifically uh apparently the
1:23:07
uh models that do product photography so men and women that uh you know wear a bunch of clothes for like an Old Navy when they're releasing a new collection like that work has been very impacted because that talent, they don't have a brand yet, right? And so maybe certain
1:23:22
And so maybe certain companies will just say like, >> "Yeah, let's just take this shirt that we have and just generate it on 20 different AI models and you're >> uh and we're good to go, right?
1:23:32
It just doesn't doesn't really matter that much if they're using real talent or not."
1:23:37
And so they choose the easier, cheaper route.
1:23:41
>> Yeah, I think that's a great example.
1:23:41
I mean, for the most part, you know, across the board, adoption is generally correlated with growth, >> but where we're seeing reductions, I mean, I think I think the creative fields are a great example.
1:23:52
So, you know, if you need like video B-roll or just like, you know, stock images or just, you know, podcast intro music, you know, you know, that that is like very easy to get from these kind of AI generated, you know, creative elements. >> Sure. >> Copyrighting.
1:24:08
So it's so fascinating because how many people Yeah.
1:24:14
It's like it's just quite interesting because when you when some of these things how many people were actually in those roles like would it actually does it show up in labor data at a at a large scale at all? Right.
1:24:28
People that are just doing you know stock photography and making their living that way or >> Yeah. Yeah.
1:24:34
And a lot of these people might have sort of slloshed around like I I mean I see uh I see Instagram reels from people who years ago were posting like After Effects tutorials, Premiere Pro, Da Vinci Resolve, like little video editing tutorials.
1:24:48
And now they're posting like AI enabled workflows and instead of showing you how to deal with a green screen the oldfashioned way, they're just doing it the new way.
1:24:58
and they're probably still doing it for clients and the client the the like the client is like spec is just like I need ads that convert and they're just doing more of the work but then there's other stuff that's bleeding out all sorts of different stuff.
1:25:12
>> I mean one one kind of framing I I I would put this in is that you know the we're seeing a lot of kind of automation of things that are very taskbased things that are like really micro jobs that aren't like full jobs at all.
1:25:23
So we're seeing like declines in freelancing across the board.
1:25:27
So freelancing is hit pretty hard, but that's really an environment where people transact. They're not >> Yeah. >> Yeah.
1:25:34
This is why we were just talking about this earlier.
1:25:36
The, you know, historically like if you needed a really specialized website, like it's not your main site, but let's say in our case, we're doing a drop.
1:25:44
Three years ago, we would have gone Yeah.
1:25:48
>> and maybe gone to Upwork and and said like, "Hey, I need a simple website made and just find somebody to do that oneoff."
1:25:54
And now AI is just so good at >> for like a basic logo for a first draft.
1:25:59
It would be like a 99 designs.
1:25:59
Before you bring in like a real branding firm, you might just get a freelancer to mock something up for you.
1:26:05
Now image generation can do that for sure.
1:26:09
>> What do you make of the computer science shifting?
1:26:11
Because there are so many opportunities for entrepreneurs. Startups are growing.
1:26:16
there's some tech layoffs, but at the same time, it feels like just in general, if you're if you have a computer science degree, you're probably going to be bit better at using AI broadly.
1:26:26
And so, there's lots of opportunity and yet uh the the number you have here is computer science enrollment is down 28% from its 2022 peak. >> Yeah. Yeah.
1:26:37
So, I have mixed feelings on it.
1:26:40
First of all, it's very dramatic. >> Yeah.
1:26:42
>> Yeah. So one thing that that kind of one optimistic take is that the supply side of labor markets is actually quite responsive to changes in technology and that wasn't obvious before and you know if people can reorient themselves flexibly >> that's great that means you know we can
1:26:58
we can be adaptive we can have more of a dynamic economy and worry less >> so I'm encouraged by that responsiveness I think it's an overreaction for two reasons one is that we are not seeing declines in employment, you know, based on the firms that are adopting a lot. And and that's true in engineering, it's
1:27:15
And and that's true in engineering, it's true in tech.
1:27:17
We're not seeing mass layoffs despite the narrative.
1:27:19
So, I think it's premature for that reason.
1:27:24
Another reason is that I think even just a couple years ago, maybe even less.
1:27:27
I mean, time is like elusive to me, but um I think you know, not so long ago, you know, we thought of AI as chat bots and code assistants. >> Yeah.
1:27:38
>> And now it's more agentic tools.
1:27:38
So it used to be such a low barrier to entry type of technology where you know anyone's grandma can use it and you know coders you know engineers were really just like you know replacing their work at high rates.
1:27:56
>> Now you know we're seeing you know complicated tools like you know aentic systems are hard to use. >> Yeah.
1:28:03
>> Yeah. they they they kind of favor the digitally native and people who have experience with engineering >> and even when they're easy to orchestrating there's there are a whole bunch of like uh from a business from an enterprise perspective like cost tradeoffs privacy security how how deep
1:28:21
is this system like even just firing up a a coding agent today you're hit with prompts like do you want this to be have access to your documents folder and that's like a question and and a lot of consumers are like I don't know and a lot of it businesses are like I don't know. So there is some sort of like uh
1:28:37
So there is some sort of like uh capability overhang. >> Yeah. Yeah.
1:28:42
And I think you know it's it's a different job than it was before you know like people are you know engineers are spending less time you know >> you know doing the front-end engineering for a website but they're doing more of kind of that devops. Mhm.
1:28:56
>> So I think it's premature and I think we'll I mean you know I suspect we might have a shortage of engineers in the way that now we have a shortage of radiologists.
1:29:05
Everyone was nervous that like radiologists were going to be a thing of the past and and now there's a shortage and you know wages are super high.
1:29:14
>> It's like the final boss of AI automation.
1:29:18
Every AI researcher is like one day I'm coming for you radiologist.
1:29:20
You imagine that it all started with like a radiologist just bullying an AI researcher and being like, "What you're doing is so useless."
1:29:28
And the AI researchers like, "I'll show you radiologist.
1:29:32
I'm going to put you out of a job."
1:29:34
And the radiologist just like, "I'd like to see you try."
1:29:35
[laughter] And then years and years go by.
1:29:37
Um, talk to me about hires to posting ratio. It's down 38. 6% since late 2022.
1:29:46
I can imagine that there's a lot of slop posts.
1:29:50
Uh we were debating this before, but yeah, how do you tease that out?
1:29:55
What do you make of the hires to posting ratio dropping?
1:30:00
>> So, this is the thing that I get the most nervous about. >> Mh.
1:30:05
>> So, you know, we're seeing some slop posts, some slop slop job postings. >> Yeah.
1:30:10
>> But we're also seeing a lot of slop applications.
1:30:12
So when a job goes up, you know, you get I don't know if you guys have posted a job recently, but I I just did last week and I got, you know, a thousand applications in the first like five minutes.
1:30:20
There are all these like job boards that are kind of helping people auto apply. >> Yeah.
1:30:25
>> Um even Indeed is doing this, which I think is a bad move for the record, but >> they'll do what they want.
1:30:29
Um it's, you know, so so basically employers are getting completely signal jammed.
1:30:34
They're getting overrun with these applications that look strong, but they really have no way of verifying.
1:30:40
>> So the utility of each job posting is going down.
1:30:42
It's not it's not as good of a way to find candidates anymore.
1:30:46
>> So employers are relying on networks.
1:30:48
It's getting harder to hire.
1:30:48
And in the economy at large, we have this kind of low hire lowire environment where there's just not a lot of movement in the economy.
1:30:56
And I think that is the result of, you know, you AI usage in the search and match process.
1:31:02
M I Yeah, you would think that uh like I I've been surprised that social media has not been that overrun with slop.
1:31:13
Like there's there's definitely some slop problems here and there, but the in general the algorithmic feeds have been sort of set up to deal with this where the bad slop gets filtered out pretty quickly >> except for LinkedIn. But yeah, >> sure.
1:31:30
Um, but I I've been I've been surprised that uh that there hasn't been as much of an intermediary where you put up a job post.
1:31:38
Yeah, you get hammered with a thousand applications, but the filtering is really really good so that you're really only looking at the top 10.
1:31:47
Maybe you dip into the top 100, but you're not at all annoyed by the bottom 900.
1:31:52
because I guarantee you that there are millions and millions of sloppy Instagram videos out there that would annoy me if I saw them, but the algorithm will just never show them to me.
1:32:01
And then maybe there's one that uses AI, but it's good and it will show it to me because I still enjoy it.
1:32:06
So I it it feels like hopefully there's people working on this.
1:32:09
I'm sure that people are, but that feels like the the next iteration to unclog this because that seems like a major problem like you need the matching in the in the uh US economy to be really really strong. Yeah.
1:32:19
I mean, there's been some regulatory challenges there, too.
1:32:23
So, a few years ago, it became illegal for employers to, you know, sift through candidates using AI. >> Wow.
1:32:30
>> And I don't know how enforced that is. >> Yeah.
1:32:33
>> But it's it's a liability for employers and not a liability for candidates.
1:32:36
So, there's some asymmetry in how in who can use AI.
1:32:41
>> That's very interesting. I had no idea.
1:32:43
When did that Yeah, I remember.
1:32:43
I remember that that you can't use AI to to to to filter out candidates.
1:32:48
I think of it as like so I I I I understand where that came from on like bias based into models and like very preliminary uh deep barely deep learning algorithms to sort of like look at the person's name and look at the graduation date and like try and filter for that.
1:33:06
Like I'm just thinking about like did the is the resume complete slop, you know, like a complete like a panggram level that doesn't seem to impose like bias in the same ways that they were trying to avoid.
1:33:18
So we're in this weird like knock-on effect world, but that's the way these things go. Jordy, anything else? >> No.
1:33:24
Come back on as there's more come back on as there's uh yeah, more more data that's notable. >> Yeah.
1:33:32
>> You can tease the hedge funds a little bit. >> Yeah. >> Give them a taste. >> Yeah, for sure.
1:33:35
>> And congrats on the progress.
1:33:35
Thanks so much for coming on.
1:33:37
>> Yeah, great to meet you, man. >> Talk to you soon. Cheers. Have a good one.
1:33:40
Let me tell you about MongoDB.
1:33:42
What's the only thing faster than the AI market?
1:33:43
Your business on MongoDB. Don't just build AI.
1:33:46
Own the data platform that powers it.
1:33:48
And let me also tell you about Cisco critical infrastructure for the AI era.
1:33:52
Unlock seamless real-time experiences and new value with Cisco.
1:33:54
Up next, we have Akos from Takeoff.
1:33:57
He's the founder and CEO.
1:33:59
He's been on the show before, [music] but this is the first time a new flag. >> How's it going? give us the news. >> Uh, it's going well.
1:34:07
It's great to see you guys.
1:34:09
Um, the news is that Sierra just bought us.
1:34:12
Uh, he announced it last Thursday.
1:34:16
>> You already completely missed that. >> Let's go. That's the first miss. That's the first miss. >> First miss. >> Brutal. Well, sorry to do that. Congratulations.
1:34:28
>> I got to go back go back to practice. I'll work on this.
1:34:32
>> So, >> no, I'm excited. I'm excited. Thank you guys.
1:34:35
So tell us the story of the company.
1:34:37
I mean we got we got enough time for I think for you to tell the entire story from start to finish because uh all of this sort of happened pretty quickly.
1:34:43
Uh where were you before you started the company?
1:34:46
When did you start the company?
1:34:47
What was the growth like?
1:34:48
Take us through the the the journey. >> Yeah, absolutely.
1:34:51
So we started the company a little over a year ago.
1:34:53
Um we started to build basically like agents that would be slightly more capable than what we're seeing today. Right.
1:35:00
I we fundamentally founded the company on the hypothesis that there were two different kinds of agents.
1:35:04
There's human in the loop agents and then there's truly autonomous agents.
1:35:08
Human in the loop agents are the agents that we all love to talk about like we're talking about cloud codecs things that you prompt.
1:35:14
They do things they can do them for a very long time.
1:35:15
It could be minutes uh dozens of minutes, hours, even some in some cases days.
1:35:20
>> But fundamentally you are the person that kicks them off and evaluates their work and then like you were talking about in the previous interview clicks accept uh viewing my downloads folder. >> Yep.
1:35:29
Autonomous agents are not that.
1:35:32
Autonomous agents when you think about it from the perspective of a buyer, it should feel like you are multiplying your labor force.
1:35:37
And when I say feel like, I mean it should be a onetoone translation.
1:35:39
I should feel like when I buy takeoff, I'm buying a 100,000 agents that can do what I might have a team of disparate humans, software, etc.
1:35:46
doing today, but at a much more massive scale.
1:35:50
>> That was like the foundational thesis of like how do we actually build agents that can do this?
1:35:52
We call them long horizon agents.
1:35:54
We call them fully autonomous agents.
1:35:55
And then we'd go after explicitly revenue aligned use cases.
1:35:59
And the reason we went after revenue aligned use cases is well there's a lot of different reasons, but the most obvious reason is like why are you going to buy missionritical AI software from a kid with crazy hair?
1:36:10
Like the only thing that's going to get you to do that is if I can prove to you I'm going to make you more money, >> right?
1:36:14
And the way I get the way I get to prove to you I'm going to make you more money is I make it very zero risk for you.
1:36:17
I'm like give me your lowest quality leads if I'm talking to a lending company.
1:36:20
Give me your patients that are going to like churn if I'm talking to a healthcare company.
1:36:23
And I'm like, let's see what I can do with the agents that I build for your company.
1:36:28
Let's see if I can recuperate that lost revenue.
1:36:29
Let's see if I can increase your top line.
1:36:31
And if we are all successful here at the end of the day, I'm going to be in your board deck at the end of the year because you bought a piece of software and revenue is up double digits.
1:36:38
Yeah, >> that was the foundational pitch, thesis, the whole idea of what the company was going to be.
1:36:42
We tried building agents in different in different ways.
1:36:44
We started actually with browser agents because we figured if we can use software, we can do what humans do.
1:36:48
But we thought that that would actually we realized not thought we realized cuz we get spit um eaten up, chewed up, spit out by the market over and over again.
1:36:55
Um we realized that that's >> over and over again for like for like 6 months [laughter] >> for 6 months.
1:37:01
For 6 months that's fair.
1:37:03
>> It's not like you were like, "Yeah, we getting chewed up one simple [laughter] >> for the viewers."
1:37:07
What Jordy and John are referring to, if you haven't read our literature, which I don't expect you to do, um is that we entered this calendar year at effectively $0 in committed revenue.
1:37:16
And by the time we got acquired by Brett and Sierra, we were at near eight figures in revenue.
1:37:21
So what they're referring to is that very short um and vertical kind of no pun intended takeoff in revenue ramp. Let's go.
1:37:28
>> Um and so >> more people should name their company Takeoff. >> Great. That's the lesson. >> It's amazing. Yeah.
1:37:34
>> Um I mean it's got its own SEO things. Former member of Migos. Rest in peace.
1:37:39
Now we're honored to like carry the name um with a positive light.
1:37:41
Um, that being said, like you're only going to make money if your agents are trying to sell or like, you know, trying to be sold upon the value of adding revenue if you actually add revenue.
1:37:53
>> And so we derisk it because I, you know, I'm not Brett Taylor.
1:37:55
I can't walk into a room or at least I couldn't walk into a room previously and get someone to pay for something that's not already driving results.
1:38:00
So we go in, we do pilots.
1:38:03
>> They're not necessarily free, but they're paid on outcome.
1:38:04
And so if I drive this outcome that we're talking about usually directly revenue or something tied to revenue for example for a lending company loans funded loans originated >> I get you're going to pay take off the way you pay a human being on commission and some sort of base units based on the amount of tokens voice SMS that are used.
1:38:22
>> So the customer thinks I'm only paying when I create x thousands of dollars in margin I'm paying hundreds of dollars of cost of goods sold.
1:38:28
They love that trade-off and they're like if it fails it fails and if it succeeds we're making more money at the end of the year.
1:38:33
Yeah, >> that's how a kid with crazy hair walks into a room and ends up selling multi-million dollar contracts over and over and over again because like once it starts actually working even what I did not expect really is like the compounding nature of of of exponential growth. >> I love it.
1:38:47
>> Walk me through how deeply you're integrating or you were integrating with some of those first customers because there's a world where you're just like give me the stale leads.
1:38:55
I will go off and I will do the email. I will do the SMS.
1:39:00
I'll do I'll do the whatever uh happens and I'll sort of like either build my build those systems or maybe you'll set up your own like Mailchimp account or whatever you want.
1:39:08
And there's another version where you're like I just I want to live within your CRM within all your tools.
1:39:13
I will do API integrations into whatever legacy systems you have to actually collect all the knowledge to make the correct move and actually drive revenue.
1:39:24
>> No, it's a fantastic question.
1:39:24
Um and it was it's categorically the latter.
1:39:26
was it's categorically the latter. So there was this there's this lecture for lack of better words I give for every potential candidate uh and existing employee of takeoff which is >> we are given the privilege right not the right but the sheer privilege of sitting between our customer and their revenue
1:39:42
like the I could explain this in a million ways why it's so important but the most important thing to explain is that our buyer was always the CEO or seuite member it wasn't some VP of something that reports into something that reports into the CEO when you are selling revenue you are selling to the CEO that's what he or is getting graded on at the end of the year. Whether
1:39:58
Whether they're a public company, of which some of our customers are, or whether they're a massively um multi-billion dollar private company, they're getting graded on revenue.
1:40:05
And then >> fired at the chief revenue officers in the audience, but no, [laughter] >> no, I know I know with you.
1:40:13
>> It's it's one of the things Brett pointed out, right?
1:40:14
Like it's kind of amazing that like every one of your customers, your contact, like the person who's in my iMessage, top five message is the CEO.
1:40:21
Um and so like to answer your question I would give this lecture that when we are referenced by our customers they have to think of us as their best employee.
1:40:30
When I say us I mean myself like Akos Spencer Shre [clears throat] my teammates names they have to think of us as their best employee.
1:40:37
And the way we get there is we have to understand their business >> as well as any individual that works for them.
1:40:42
They could be a 10,000 employee company.
1:40:44
They should be able to ask us about anything that is even remotely related to the line of work that our agents are doing for their business and we should be able to answer it.
1:40:50
I'm talking about gross margins.
1:40:51
I'm talking about conversion rates.
1:40:52
I'm talking about um time to fund.
1:40:54
I'm talking about time between first contact to revenue generated. Literally everything.
1:40:57
And we know we've succeeded when the CEO starts asking us questions about their business.
1:41:03
That's when you're in like, you know, the promised land.
1:41:05
That's when you are literally their friend.
1:41:06
When they're texting you 5:30 in the morning, 10:00 at night, um, and like none of your family or friends are in your top five message anymore.
1:41:12
It's just your customer CEOs.
1:41:14
And so it's very much understanding like the intricacies of that business in order to build an agent that could actually do what's going to drive that company's revenue.
1:41:22
And so we have to understand every every piece of software that they're using, every single thing that somebody might do because we are trying to genuinely scale the workforce.
1:41:30
And you can only scale the workforce if you can do it end to end.
1:41:34
And that's like a really important thing that I think most agent companies don't get.
1:41:36
If you're going to sell an agent into some work stream, but you're only going to take like a horizontal slice, it's virtually useless because then you have to educate everything below and above it how to drive the end to end result. >> Mhm.
1:41:48
>> So, if you want to actually drive the business outcomes, own the whole thing.
1:41:50
If you want to own the whole thing, you have to be capable of owning the whole thing, which means you have to understand the business well enough to build the agent to do so.
1:41:55
You guys had Marky, who's been a friend and an incredibly uh incredible founder who I've learned a lot from, on the show, I think a month or two ago.
1:42:01
And Marky talked about how her entire company and product is rooted in this foundational philosophy that we have to translate whatever language the company is speaking to what the agent is going to do. >> Yeah.
1:42:13
>> Like we think very very similarly.
1:42:13
Our culture is entirely predicated on that assumption that if we don't understand the business better than our customer or as well as our customer, our agents aren't going to do it as well as we need them to.
1:42:22
So, what does that translation look like for you?
1:42:24
Is it uh a bunch of markdown files and then your agent can interpret those?
1:42:30
Because you could go all the way to like we pre-trained a model just for you and then we could be like we fine-tuned a model for you sort of the thinking machines model and you could be like well we're using you know the frontier models but we have a custom harness for you or we customize our harness for you or we write a special integration or it's just or the agent just shows up and it figures it out.
1:42:51
I love that you're giving me multiple choice because if you didn't I would just like ramble.
1:42:53
Um it is the second half of answers that you just said.
1:42:55
So like another foundational philosophy that we kind of built this company on is that the inference API is a commodity. That's a sound bite.
1:43:02
You can clip me that'd be great.
1:43:03
Um [laughter] and that's a crazy thing to say, right?
1:43:04
It's a crazy thing to say that an inference API is a commodity because we think about claude and anthropic sorry anthropic and open AI at tens of billions in revenue.
1:43:12
People are going to be like that's all inference. I would disagree.
1:43:15
I would say it's a function of the things built on top of inference.
1:43:18
We're talking about claude codeex. interesting claw code. >> Yeah.
1:43:22
>> And what we need to be able to do is you can call these and most people would call these harnesses, right?
1:43:27
Like harnesses with like a great GUI with a great command line interface, >> but it's functionally the thing that's delivering endto-end value >> and like coding was a great first coding and chat agents were a great first product because that was the endto-end value.
1:43:40
Again, it's a human in the loop type agent.
1:43:41
So, you're giving the value to the person that's using the product. >> Yeah.
1:43:44
The second the subsequent type of like wins in enterprise AI and when I say wins I don't mean like hundreds of millions in revenue or even billions.
1:43:51
talking about like the next wave of tens of billions of revenue.
1:43:54
>> It's going to come from the fact that your harness, which is just a fancy way of saying agents that can do multiple things and operate across multiple different services surfaces as opposed to a single column response API should be as capable as someone that is like is on a job listing or like something that you are hiring to drive an outcome or a result for the business.
1:44:12
And so it's a combination of harnesses and specifically at takeoff what we built was what we call it as a DSL a domain specific language, right?
1:44:19
So you should be able to educate, direct and build the agents on takeoff using our domain specific language that is built around the idea of we're trying to handle something end to end.
1:44:31
>> The other kind of like unique thing about this is like your agents have to be answerable to the outside world.
1:44:34
Like you can't just say go do a thing.
1:44:38
>> The thing that you are that our customers are calling APIs for is go fund this loan or an API call to go onboard this patient or go get this patient's prior authorization.
1:44:46
that requires multiple actions by the agent that then in turn require input from the outside world.
1:44:52
Let's use the borrower example, the loan borrower.
1:44:54
>> If you're getting an API call to your agent that says go fund this loan for this borrower, this lead, you have to call that lead, contact them, help them with that initial like rate quoting, understanding what options they have, whether it's a heliloc, a refi, a home equity loan.
1:45:04
Then you have to have a second call after whatever happened between the first call and the second call.
1:45:08
You have to go reach out to third parties, the e- notaries, the underwriters, everything else involved, document collection.
1:45:13
Then you have to have a third call saying, "Hey, I noticed you got stuck here because you have this like weird Iowa borrower question about co-borrower um co-signing."
1:45:21
Then you have a fourth call.
1:45:22
It's getting it over the line for funded.
1:45:23
This is something that for an agent to actually handle end to end, >> your harness is like is is transcending just like tool calls, >> right?
1:45:31
It's like it's an always on harness with a heartbeat that's answering anything that could happen by on behalf of or in relation to this like central entity in this example of the borrower.
1:45:41
Tell me a little bit about post merger integration.
1:45:43
I could see Sierra having a product called take off.
1:45:47
I could see Sierra just being a service or company that you work with for a bunch of different things.
1:45:54
And I don't even know if I need different products because AI is so broad that everything sort of merges together into like one product that can do multiple things and I just flip on a switch and say, "Okay, I want you to handle this.
1:46:05
I want you to handle this."
1:46:06
Uh, but how are you thinking about integration?
1:46:07
I know it's I know it's like really really early, but uh I imagine that this was what you were talking to Brett about was like a vision of what these two companies can do together.
1:46:17
So like take us through a little bit of it.
1:46:19
>> Well, I mean that's actually I'm going to go in reverse order of your questions here.
1:46:22
Like when when Brett and I first chatted uh we basically realized that we have a similar vision of what the world was headed towards.
1:46:29
And what we realized was that by virtue of just again being a kid with crazy hair, like there's no chance that I was going to like compete and win in customer support.
1:46:36
There's a dozen companies, three of which [laughter] that are like >> selling yourself short.
1:46:42
You you seem you seem you seem to me I'm I'm getting young Brett Taylor. >> Yeah.
1:46:49
[laughter] Let's pull up a picture of Brett Taylor's hair, please, and see if you can hair.
1:46:54
>> I think you [laughter] got a I think you got a shot. But yes. Okay.
1:46:57
>> The point being that like we had to come from a different angle, right?
1:46:58
We had to sell a thing that only the early adopters were ready for >> and like like it's not like we and like when when I say we had a similar vision of of the direction the world was headed in like we were further along on that timeline and like we had done this thing that I don't think most of the world realized was possible yet and it wasn't until we proved it was right.
1:47:15
That's what was really exciting to I think Brett and the company is like, hey, we would love to get to where you are, but we're realizing that you're already there and >> why not get there together and then scale it times a million. Sure. Right.
1:47:27
And so that's kind of how the original conversation started.
1:47:30
>> Um we then kind of came to this like realization that and I think what was actually really interesting for you guys to understand or for anyone who's listening and watching is that we realized we were on to something when our first like three seven figure customers were like they already had customer support vendors, right?
1:47:44
they had like a Sierra or a Decadon or something else there.
1:47:48
>> They were spending on average between a few hundred grand to maybe like maybe a million dollars with them.
1:47:52
With us, they were spending at least three times more. >> Wow. >> Right.
1:47:58
So like they had they had an AI support vendor and they also had takeoff.
1:48:01
They were spending three times in in one case eight times more >> than they were spending with their support vendor.
1:48:06
>> And that makes sense because you're driving revenue and you're talking >> and if you're driving revenue there's three kinds of software, right?
1:48:10
There's revenue driving software, there is functional software and then there's must have software.
1:48:13
And if you're the first category, Google ads, Facebook ads, $1 in equals more than $1. I will keep flat.
1:48:20
>> And that's what we were going for.
1:48:20
It's exactly how we want to be thought by our customers.
1:48:22
And so we realized, you know, again, all the things that Brett and I were talking about that we were excited about, a lot of the same simil shared ideas around where the world was headed.
1:48:30
>> Um, we're like, hey, together this can be 1 plus 1,000.
1:48:33
>> And so we uh we I don't know if you guys saw, probably not because you have a lot going in your mind.
1:48:36
We uh we as in Sierra and take off Sierra together launched uh this product called Horizon which is this new thing and the reason it has to be this net new thing is because we want people to realize this is a step function jump in capability. >> Yeah.
1:48:51
>> A step function jump in capability which is going to drive revenue for your business.
1:48:54
It's not just agent and like one second [music] cost savings but it's agents that your CEO is buying.
1:48:58
And that's really freaking exciting. Yeah.
1:49:02
>> I don't know if I can swear. I'm sorry.
1:49:04
Um but that's really exciting.
1:49:06
>> [laughter] >> That's awesome. Yeah.
1:49:06
No, it makes it makes so much sense.
1:49:08
>> You're great at You're great at naming. >> Yeah.
1:49:10
Yeah, these are all every every name is great. Like these are all good. Yeah. >> I love them.
1:49:14
Uh well, thank you so much, Jordy. Anything? It >> was great. Great to meet you.
1:49:17
Uh I found the I found the the whole pitch very compelling.
1:49:21
I was just imagining myself as a as a CEO or enterprise buyer just being like, I'm sold. Just send the contract. Send the contract.
1:49:29
>> I will say like I've been on a few sales calls with Brett now and Brett and it's very flattering to hear this from Brett Taylor, right?
1:49:33
one of the best salesman probably ever lived in software's history.
1:49:37
>> We've had a few sales calls together.
1:49:39
>> It is magic in that room.
1:49:39
Like people get really freaking excited when we show them Horizon and it's like it like people start imagining what they're going to be doing for their business, all the awards they're going to get.
1:49:47
The fact that in the board deck it's going to be plus double digit percentages at the end of the year and that is very exciting for us as a company. >> It's very exciting. >> Amazing. >> Thank you so much.
1:49:56
>> I know I can see why you guys did the deal. Great. Great to hang dude.
1:49:59
>> We'll talk to you soon. We'll talk soon.
1:50:01
Let me tell you about public investing for those who take it seriously.
1:50:04
They got stocks, options, bonds, crypto, treasuries, and more with great customer service.
1:50:10
>> Mark Zuckerberg is in the Wall Street Journal opinion section with a new piece, The AI future is for everyone.
1:50:14
He says, "The history of democracy and economics has proved that centralized power stifles human potential." And uh it's quite long.
1:50:25
G let you guys read it, but let's head into the comment section.
1:50:31
Let's get a quick Let's get a quick rea Let's get a quick reaction.
1:50:35
>> Uh this is the Wall Street Journal.
1:50:35
I think it'll be >> um No, it it it looks relatively tame.
1:50:39
It'll be >> um but yeah in yeah you making a you know a clear effort to position to to be the overtly >> there was a white space for a guy investing hundreds of billions of dollars a year in AI that is like says hey this is going to be really great for everyone.
1:50:58
>> Yeah >> and I'm gonna help us get there.
1:50:59
>> and I'm gonna help us get there. It is it is interesting like Facebook does have some monopolies but like the competition for attention is constant and there are always sources outside like they've never had a full monopoly on uh on social media even with Tik Tok and Snapchat and uh LinkedIn and Twitch and YouTube and Netflix and the podcast feed and SMS and iMessage like there are
1:51:32
so many other platforms for disseminating information like I I don't know I I it's hard to jump straight to a critique here but um the the key quote that Andrew Curran pulled out was that he said in most cases like cyber security the history of open source software has shown that giving everyone full access to powerful systems will be the best way to protect safety and security over time. So, he's firmly on
1:51:56
So, he's firmly on the side of uh democratizing powerful AI and uh he is yet another one.
1:52:02
Uh I imagine that they that they signed the letter.
1:52:06
I I've lost track at this point, but you can imagine that he did.
1:52:08
Um anyway, thank you so much for tuning in.
1:52:14
The other piece of news is that Apple is launching Apple Upgrade next week, an iPhone, iPad, Mac, and Apple Watch leasing/subscription program.
1:52:24
Uh they said you will own >> and you will be happy.
1:52:28
[laughter] >> We're launching our new program.
1:52:30
You will own nothing and be happy.
1:52:33
>> Uh [clears throat] it's partnering with Clara to launch in the United States at online and retail stores. Uh it's now official.
1:52:37
Leasing prices start as low as $20 or $17. 99 per month for iPhone. Uh $11. 99 for Apple Watch, $24. 99 for Mac, and $11. 99 for iPad. So uh interesting.
1:52:51
I mean, a lot of people are saying this is a direct reaction to uh increased prices for memory, increased prices for products.
1:53:01
There was a time when an iPhone was a couple hundred and there were incentives to jump on a Verizon plan and you sort of advertise the cost over that. Those days are gone.
1:53:11
Like we're in the world of like a $2,000 iPhone.
1:53:12
It's a significant >> Same thing with our gongs for people.
1:53:17
>> So, you want a subscription gong?
1:53:18
>> No, I'm just saying there was a time when a TVPN gong was $200.
1:53:20
Now it's in the tens of thousands of dollars, right?
1:53:26
>> It's actually so expensive.
1:53:26
Somebody a friend of mine texted me and was like, "Where do we get the gongs? I need a gong."
1:53:30
And I was like, "I think you should start small."
1:53:32
And this is not like you're you can't handle the big gong.
1:53:35
I was more saying that like there is a joy to being on the hideonic treadmill of larger gongs.
1:53:40
Like you don't want to jump straight to the biggest gong.
1:53:43
You want to start with a small gong >> and work your way up. >> Work your way up.
1:53:46
Because every gong that we've added has been so electric.
1:53:51
when we get >> I think it's time for a new one.
1:53:52
>> You want an even bigger gong or what?
1:53:54
>> Yeah, I want I want one that's hanging from the rafters.
1:53:57
>> You want a giant gong? Maybe.
1:53:57
Um but and also every gong has a different flavor, different sound, you know, different amount. You got to warm them up. All sorts of things.
1:54:06
Uh we always warm up the gong.
1:54:11
>> Anyways, folks, that's our show for today.
1:54:13
Enjoy the rest >> of your July 28th.
1:54:15
Leave us five stars on Apple Podcast and Spotify. >> Money never sleeps. You shouldn't either. Call me back.
1:54:25
>> Sign up for the newsletter at tvpn.
1:54:25
com and we will see you tomorrow at 11:00 a. m. Pacific. Goodbye. >> Cheers. >> Flashback. [screaming]