0:07
[Music] Hey, hey, hey.
Search this video
0:29
Hey, [Music] [Music] do you [Music] [Music] Hey, [Music] hey, hey.
1:16
[Music] [Music] Hey, [Music] hey, hey. Hey everybody.
1:48
[Music] [Music] [Music] All right. [Music] [Music] Heat.
2:42
[Music] [Music] Hey, Heat.
3:03
[Music] We came to this world.
3:27
We came to this world to shape a future.
3:38
We came to this world [Music] to reach the stars to shape our future.
3:52
[Music] Feel [Music] the music. [Music] Feel that music.
4:54
[Music] You're watching TVPN.
5:03
>> Today is Thursday, September 25th, 2025.
5:06
We are live from the TBPN Ultradome, the Temple of Technology, the fortress of finance, the capital of capital.
5:10
Good morning to everyone in the chat.
5:13
Good morning to Gold Rock, Reggov, John Techno Chief of course.
5:18
Uh the post of the day is Alex Heath in sources. I printed it. He says scoops. He's scooping.
5:29
>> He says OpenAI is planning to bring ads to chat GPT.
5:32
Also in a message to employees, Sam Alman says he wants 250 gawatts of compute by 2033.
5:39
He calls OpenAI's team behind Stargate a core bet like research and robotics.
5:46
Doing this right will cost trillions, he says.
5:52
Well, hopefully he can save time and money with ramp. com. Time is money save both.
5:56
Easy use corporate cards, bill payments, accounting, and a whole lot more all in one place. >> We go.
6:01
But I was thinking about uh what Sam Alman's projections have to say about um the the the the the buildout of AI.
6:11
We're in pretty crazy times.
6:16
Certainly, lots of people are comping it to 1999.
6:17
Lots of people are saying 1996 if they're bullish.
6:21
Uh the exact year matters a lot.
6:24
Uh I mean, even 1999, I feel like there was still >> some people are saying this time is different. >> 18 months, you know.
6:29
Um but uh you know the technology is even more magical than the internet potentially and of course the internet already exists so it's easier to scale revenues as we've seen.
6:43
Um and so my my take on this was uh the AI deals are getting crazier.
6:47
Obviously this week we saw Nvidia is going to invest a hundred billion dollars in OpenAI which in turn will be used to buy Nvidia GPUs.
6:55
But we aren't fully in 1999 mode.
6:59
Even though a lot of people are saying that because big tech just has so much cash to fund the buildout.
7:03
And that wasn't quite what was happening in the dot boom.
7:07
In the dot boom there was a lot more leverage.
7:09
The price to earnings ratios were crazy 10x higher on median.
7:15
And uh and so we're not in the crazy crazy debt era.
7:18
There's a little bit of that going on with Oracle, >> but it's not huge.
7:22
the debt to debt to equity ratios aren't insane yet. >> Yeah. Yeah.
7:27
Certainly players like Meta have tapped the debt markets, right?
7:32
They did an offering with with Blue Owl, but it is a >> uh small, you know, it's a it's a >> some uh minority percentage of their earnings. >> Exactly. Exactly. Yeah. Yeah.
7:42
It's >> Oracle went out this week as well for uh a $15 billion uh >> convertible debt offering. >> Yeah. Offering.
7:50
I'm sure that'll get filled quickly again because it's it's a >> uh uh feels relatively low risk even though even though they're they're forecasting well beyond what a traditional public company would be doing. >> Yep.
8:04
Uh and so but the the nature of the deals is getting more unique. It's different.
8:12
I mean we we're seeing letters of intent now uh which is just a very fuzzy >> or investments being announced that are still in discussion. >> Exactly. Right. Exactly.
8:22
This Nvidia OpenAI investment, if you read the print at the bottom of the >> It was a non-binding letter of intent. >> Yeah.
8:30
They're basically saying, "We're talking about this." Exactly.
8:33
>> This is like the the paperwork for the deal.
8:35
>> Usually, you go the opposite direction.
8:37
You get the leaks, then you get the actual announcement once the deal is fully signed, the deal signs, and then the press tour starts, but now we're putting the press tour a little bit earlier.
8:46
Um, and I keep coming back to YC Demo Day because at YC Demo Day, the letter of intent is often pulled out for companies that don't quite have the ARR or the DAUs to justify whatever they're trying to fund raise.
9:01
Um, and I wanted to revisit uh a story from YC Demo Day.
9:08
My favorite YC letter of intent story.
9:11
Uh, probably Boom Supersonic.
9:11
uh extremely we've had uh Blake on the show multiple times.
9:15
Um an extremely ambitious project to bring back commercial supersonic aviation basically bring back the Concord.
9:21
Um Boom was one of the first hard tech companies to go through YC.
9:26
There were a few at the time.
9:28
This was the Sam Alman era again.
9:30
Uh and they didn't really fit the current mold.
9:32
They couldn't show a graph of DAUs or ARR obviously because they were they hadn't even built planes yet.
9:38
They were building like plans to build planes.
9:39
There's only so much you can do in three months. It was a founder bet.
9:42
>> Yeah, it was a founder bet.
9:42
And Blake also >> guy was a product manager at Groupon. Exactly.
9:46
He's got what it takes to take us back to hypersonic. Yeah.
9:48
So, it wasn't >> It sounds It sounds like I'm joking, but I'm totally not.
9:54
When you see Blake's level of conviction and uh his inability to >> not go supersonic. >> Yeah.
10:03
But it wasn't like he was like former locked and was spinning out some technology that immediately could just like piggy back on the shoulders of giants immedately sp Yeah, exactly.
10:12
He had to really build stuff from scratch and put the puzzle pieces together and one of those puzzle pieces was a massive LOI.
10:22
And so he stood on stage and said, you know, you're going to think this is a sci-fi project.
10:27
you think this is a moonshot project, but the one thing that I can tell you is that I have an LOI.
10:33
It's nonbinding, but it's for $2 billion.
10:35
And so that was what went on the screen, I think, around there.
10:38
And he had gone to Virgin, talked to probably Richard Branson, and Virgin had optioned 10 airplanes in a non-binding LOI and the deal value was roughly 2 billion.
10:52
And so if if it was fully exercised now I think people generally believe this is a kind of business it's like if you build it there will be demand. Exactly.
11:01
There will be demand from the airlines.
11:03
There will be demand from consumers.
11:06
>> But having the LOI means something it shows that you can go in that room and convince them that if you build it they'll buy it.
11:13
And then you and then that leads to investors being saying okay well I will back you.
11:17
Um and so booms had ups and downs. They had a down round.
11:22
A lot of YC folks came back in, invested in the company again just to keep it going. And Blake's still at it.
11:29
Uh he's getting regulatory approvals, which we saw, and building planes one at a time. He's actually flying.
11:32
Um and as loose as non-binding LOIs can be, uh they still help pull forward the future and can be a useful tool in making everyone around the table feel like there's mutual buyin because everyone's at least said, "If this happens, I'm in. If that happens, I'm in.
11:48
And if everyone commits and if the capital partners commit and the customers commit and everyone commits then the thing should actually happen.
11:57
And so um the boom story for me underscores how unclear the path of technological development can be.
12:03
Sam Alman is now guiding toward 250 gawatts of compute at open AI by 2033 which is a staggering number.
12:11
We saw a post which we'll get into.
12:12
A single gigawatt data center is many football fields long visible from space.
12:16
It basically terraforms the earth.
12:18
Um, and while many people are writing off Sam's projection as merely attempting to win the contest of saying the biggest number, his projections aren't entirely new to the world of AI compute forecasting.
12:29
Leopold ashen Brener predicted a single training run cluster around 100 gigawatts coming online by 2030.
12:35
Uh, so three years earlier, one-third the size.
12:38
Uh, just two years after.
12:41
And then Leopold predicted in 2028 that there would be the first 10 gigawatt cluster.
12:45
um 250 gawatts by 2033 lines up with that.
12:50
But wow, America will look dramatically different if that comes to fruition.
12:54
Um in 2023, for reference, two years ago, the US was producing an average power in total for everything, not just AI, of 485 gawatt.
13:08
And so Sam is saying that just open AI if everything else stayed the same would be onethird of all power consumption in America for that entire year in 2033.
13:19
Now obviously everything else is growing and there are other labs and you would imagine that Amazon and Anthropic keep up and Google and Deepmind keep up and so um it will be an entirely new chapter of the economic story.
13:32
>> What are some of Elon's uh promises again or or or goals?
13:35
Yeah, I was talking to Tyler about this because I so I I I mapped out the power projections.
13:42
This is uh straight lines on a log graph.
13:45
Of course, uh Leupold is the taller bars here. Jordy, you can see this.
13:49
And Leopold is projecting uh logarithmic growth, which is exponential growth.
13:53
This is uh every tick on here is an order of magnitude bigger uh for gigawatts.
13:58
And uh and you can see Leopold is is projecting a straight line on a log graph basically exponential growth continuously uh based on you know how he's mapping out the few dates that he gave.
14:12
He didn't give projections for every single year.
14:14
Uh Sam is actually it's so funny because he's the Sam story is framed as like 250 gawatts by 2033. This is so insane.
14:23
But he's actually short of sort of guiding towards deceleration.
14:27
Like you can see Sam's Sam's darker bars here. >> He's being humble.
14:32
>> He's being humble relative.
14:33
>> He doesn't want to he doesn't want to, you know, >> over relative to some other folks in in the category.
14:37
So, uh, Tyler, what Yeah. What did you find?
14:39
Did you see anything from the other labs? >> Yeah.
14:41
So, I was trying to figure out like what all the other labs are doing, like what they look like on that graph.
14:46
Um, so none of them have have done real kind of projections like no one has said like, okay, by 2035 we're gonna have 250 gawatts, anything like that.
14:54
Um so so for Meta Zuck the big number that you could use for this is his um at the at the White House dinner 600 billion through 2028. >> Yep.
15:05
>> Um and so >> so it's an aggregate number. >> Yeah.
15:08
>> But you have but like they're only doing 50 or 60 now.
15:09
And so to get to 2028 that's only 3 years away.
15:13
So you have to do 50 100 200 400 something like that.
15:19
Like it has to ramp really fast. >> Yeah.
15:21
>> Yeah. I was trying to figure out like what does that actually mean in like gigawatts like is the actual like price that you could like how many could you get out of >> convert the dollars to >> and and so basically I used um I I looked at Colossus 2 which is
15:34
>> maybe the one gawatt one >> um and then estimates are this is like very like nonaccurate this is totally guessed but like maybe somewhere around 15 billion >> okay to build that >> order of magnitude 10 billion something 10 to 20 billion let's keep it super vague for a one gigawatt. >> Yeah. >> Yeah.
15:52
>> And so >> so with 600 um billion you can do like maybe like around >> 60 gigawattsif 50 gigawatts. >> Yeah.
15:59
Like something in that >> if you were to just copy and paste a bunch. >> Yeah.
16:02
And but that's I mean that's like Elon like the most efficient like this is the fastest of all time.
16:08
Like >> it took them like six months to get to to 300 megawws which is like >> double the next fastest build out which was the um Crusoe OpenAI one. >> Yeah.
16:18
So I think it's probably that's a little aggressive >> and and Zach's number was at the end of 2028.
16:25
>> Yeah, it was through 2028 I believe.
16:27
>> Yeah, I mean the the end of 2028 for for Sam and Liupold.
16:29
They're more thinking around 10 gawatts I think.
16:31
So uh interesting to see uh if if Zuck's projections kind of outpace what's there. I don't know.
16:41
And you guys saw today Core Weave and OpenAI expanded their contract to a total contract value of 22. 4 billion. >> 22. 4. Okay. >> 6 and a half billion.
16:52
So >> it's a deal for >> spraying and praying. >> Yeah.
16:57
>> Another thing you could use uh besides Leopold is the AI 2027 numbers.
17:01
>> Yeah, I was wondering about those.
17:03
>> So if you look at those like once like in once you get to 2027 after that the numbers like explode.
17:07
It's like oh yeah the fast takeoff.
17:10
>> It's like 30 trillion on on data centers. Yeah.
17:12
>> Um but but through 2026 2027 it's like around like 500 billion a year >> and that's like 500 I think it's like 650 in kind of the end of 2027.
17:22
>> So like you could imagine we're like reasonably on track for that. Yeah.
17:25
>> Considering Meta is alone is doing around 600 over >> Yeah. >> through. Yeah.
17:30
If you aggregate all the labs. >> Yeah.
17:32
So I think that's like fairly reasonable. So >> yeah it's hard.
17:35
I mean people are like if you look at the timeline everyone is like okay we're at the top like this is so bearish. Yep.
17:41
>> But if you look at a lot of people who are like super bullish on AI, we're still broadly on track.
17:43
I would say >> yeah, this is what uh what run was saying how like everyone all the richest people have voted and they've all said we're going all in.
17:52
We're putting all the chips on the table.
17:55
>> It's not about immediate ROI. >> Yeah.
17:57
And I mean honestly like if it's funded with cash and equity and not too much debt, there's a world where like you take the hit of maybe some stagnation and you get through and you are more of like the Amazon story than the next star story or I forget what I always forget the names of the companies that went away.
18:17
>> Even in the bear case it's like um there's the the book um boom by burn Hobart. Yeah.
18:21
which is like bubbles are good because uh broadly like they don't affect like the normal economy because it's like all the it's like >> the most risk on. Yes.
18:29
Uh the interesting takeaway from boom was that uh the uh the the the common narrative is that like retail gets hurt the most in a bubble but in fact it is not retail.
18:41
it is the the the the wealthiest that get their the the the biggest haircuts because they are the most leveraged at the peak of the bubble and they have the most wealth to compress.
18:51
Whereas uh the the average American through the number of bubbles that Bern Hobart analyzed uh is not super levered long the most riskon asset at the time of the top.
19:04
Uh Jordy, what were you gonna say?
19:06
>> Uh so there's a company called Iron. Oh, yes.
19:08
the company pivoting from Bitcoin mining to being an AI data center company. Yep.
19:14
>> Their stock is up tremendously.
19:14
Andrew Wilkinson uh hit the timeline a couple months ago at this point.
19:19
I think it basically like said >> uh I read it as financial advice, but he was basically making his case.
19:25
The stock is up massively since he came out and >> um made the case for for why he thought they were undervalued.
19:32
He's the current uh share price is $48.
19:35
It's up uh let's see exactly it's up from around like I guess 171 18 I think when he recommended it. So it's up massively.
19:46
He's now calling for another 7x >> because he says Iron >> uh is about to control more power than the Hoover Dam uh 2,910 megawatt. 2. 9 gigawatts.
20:00
Um and uh so anyways, he's now calling uh for the company to uh yeah, again 7 7x again he thinks that $300 per share could be a layup.
20:17
>> He lays out a potential bare case, right? >> Okay.
20:19
Um given that uh he says if nobody want if AI turns out to be a nothing burger and nobody wants iron's power land data centers and servers that could be bad.
20:29
Uh unable to if they can't really capitalize on the demand because they just don't have they're not the data centers aren't aren't configured to to the level uh that they need to be to be uh great assets for either hyperscalers or other AI companies.
20:43
So, uh, lays out some some potential downside scenarios, but, um, I I do wonder I do wonder if we'll see, um, some of these companies that are trying to take a crack at actually competing with the hyperscalers as like cloud, you know, serious cloud providers.
21:01
I do wonder if some of these companies will get to the point where they have a lot of energy that they basically just have to sell to another more sophisticated player that can um actually take over and and turn it into, you know, help it realize its potential. >> Yeah.
21:17
Uh the chat is asking for an update on TBPN merch if it will be available for our most loyal soldiers in the chat.
21:26
>> So, we got to set something up.
21:28
>> Yeah, we got to we got to figure out how to how to make this happen.
21:29
We uh the first run of merch, we didn't we didn't want to sell it.
21:33
We had never made merch as a company before and it didn't feel right to sell something that is not our core competency, but uh we had a meeting yesterday talking about uh the next run of merch and uh getting that available by the beginning of this coming year. So uh stay tuned.
21:49
Um and uh thank you for your patience.
21:52
>> We got to get Reream on some merch.
21:52
One live stream 30 plus destinations.
21:54
They make our stream possible every day.
21:58
multiream and reach your audience wherever they are.
21:59
Um, I had a friend that asked for some advice.
22:03
If you're making money during the bull run, what should you do if you feel like you're at the top?
22:11
What's the safest portfolio for a young person these days? >> Rotate into watches. >> Watches.
22:19
>> Watches. go entirely into Swiss >> I mean technology >> like if your com if your if your comparison is NFTs like yes like the watch market did actually draw down post NFT crash >> and there were people that converted NFT
22:34
earnings into watches and they probably did better >> I feel like the what was it the Rolex market sold off by like 30% or something but I mean a lot of NFTs went to zero >> still above retail in for most watches >> um but yeah I mean it feels like the
22:49
like there is a there is there's the existence of a risk curve and the the the companies that are uh trading at the highest you know you just have to go back to the basics of the Warren Buffett stuff of like company with no earnings
23:02
trading at tens of billions of dollars that's going to be the most dangerous they will be most punished in in a down economy whereas a hyperscaler that still has enterprise licenses and huge ads business is going to fare better even if they get a haircut. We saw this with
23:16
We saw this with Meta during the last cycle.
23:18
I mean, they traded down a huge amount, like almost 50% off the metaverse highs, uh, but built back up pretty quickly on the next wave.
23:27
Um, and then of course you, you know, you got gold, you got Bitcoin, you got real estate, you got plenty of things, but, um, what, uh, what would you counsel someone uh, in their 20s making decent money to do with, uh, you know, excess earnings at this point in time? I don't know.
23:46
I I think that the the the young millennials, Gen Z, are overly obsessed with investing.
23:55
>> Uh it's all that a lot of people can think about and it ends up being a massive distraction from just getting better at your craft, your career, and just increasing your your earning power. Yep.
24:05
>> So, I think the number one I mean, I talked to the guys on the team about this.
24:08
Uh we're obviously talking about uh different private companies, different investments all day long on the show.
24:14
And so it's easy to get kind of fixated and and get kind of FOMO around this stuff, but anybody knows if you've made a sizable investment or or uh you know uh if you've made an investment ever that's like doubledigit percentage of your net worth and it's trading liquid, >> uh that can be just such an >> emotional roller coaster. >> Yeah.
24:33
just such a massive distraction from what uh what you should be focused on uh in your 20s, which is just increasing your skill set, building your building your network, like figuring out how you can increase cash flow. So, >> yep. >> I don't know.
24:49
You had a you had a I think >> yeah my my my take was um there's huge alpha in not even picking the correct asset but just setting up your life so that the money actually gets into the asset before you can spend it. Yeah.
25:04
And so, uh, when I started making money in my 20s, uh, every time I would get paid, I would go to the bank, I would take out something like $1,000 of cash and put it in a safety deposit box.
25:17
And that just built up over time.
25:20
And I realized that if I was out on the weekends, I could not access it because you'd have to physically go to the bank to get the cash.
25:28
And so I I realized that if you're out with friends and they're like, "Oh, like you know, should we, you know, spend all this money on whatever?"
25:34
Um, if you have it in your checking account, you're like, "Yeah, no problem."
25:39
If you have it in your savings account, it's like supposed to be in savings, but you can just click one button and move it over.
25:45
>> I remember being a like a a teenager and realizing that your savings account was like literally one transfer away from checking. It's >> ridiculous. >> Wow.
25:54
This takes a lot of willpower to actually make it savings versus spendings. >> Yeah. Exactly. Exactly.
26:00
So, if you if you build up a literal horde of treasure, a it's incredible psychologically because you you open up that safety deposit box and you can physically see the wealth growing, which is incredibly enriching.
26:14
It's incredibly satisfying.
26:16
Uh and it makes investing very much addicting.
26:19
But even just going into your payroll provider and route like usually you can route your paycheck to a some sort of investment platform maybe public.
26:28
com investing for those who take it seriously.
26:30
They got multiasset investing interesting yields trusted by millions.
26:34
Uh and then you can set up like automated buys on ETFs if you like the MAG7 if you like gold if you like Bitcoin like whatever you want.
26:40
You can you can have automatic investments happening so that you're not spending all your time trying to become like a hedge fund unless that's actually the service that you would have liked is somebody you you go and you give them cash. >> Yeah.
26:54
>> And uh they even they use uh movie cash to give you show you the pile but they take that they actually invest >> invest it.
27:01
That's a good service but to access it you have to >> you can go look at the cash you can you can >> the physical key was really really great.
27:10
Uh it was it was >> I've never had one of those. >> Yeah.
27:13
Safety deposit box is underrated. >> Underrated.
27:15
Uh >> do they still make those? >> They do. Yeah.
27:17
Available pretty much every bank have little like a mouse you have little.
27:21
It is it is somewhat rodent coded to just be hoarding dragon coded potentially.
27:28
>> Uh you know to hoard your pile of gold >> or or could kind of cool and mysterious if it's all over the country.
27:32
You know like you've got a little stash in Dallas, you know, a little keep some in the bay. >> Yeah.
27:39
Uh, what do you think, Tyler?
27:41
>> I I actually know a lot of friends um who have like a a Bitcoin like wallet and then they have um the the seed phrase in like different safety deposit boxes all over like the state. Yep.
27:51
>> So, they have to drive it's like four hours to drive to every single bank to >> That's diamond hands, baby.
27:55
They're not they're not pay for handing anything.
27:58
>> And so, if they're just continuing to contribute to that, uh the the trick is that the cash flow typically comes in like monthly or every other week installments.
28:06
And so, you know, how do you actually set up that so that it's happening on a regular basis, but still inaccessible?
28:11
That's a little bit tricky.
28:13
Um, maybe someone can build it.
28:15
Maybe someone can use Privy, our wallet infrastructure partner.
28:20
Wallet infrastructure for every bank.
28:22
Privy makes it easy to build on crypto rails, securely spin up white label wallets, sign transactions, integrate onchain infrastructure, all through one simple API.
28:28
Barely AI says, "Looks like Nvidia and OpenAI will be paying zero investment banking fees for their $100 billion partnership."
28:39
>> Some hit me with the so sad for the investment bankers.
28:43
Uh we talked to carried no interest.
28:46
He was telling us >> Wall Street in cut Wall Street in >> Come on, break us off.
28:49
Break off Morgan Stanley.
28:51
Make break off Goldman Sachs. Those guys need a break. really it's really imp.
28:55
I mean it just these founders founder deals at this scale is is very cool. >> Yeah.
28:59
Uh I mean that's what everything is.
29:01
We we talked about this with Nvidia uh like spending like a few million dollars lobbying in Washington but Jensen's there and founder to president founder to founder.
29:11
Like that's the nature of these deals.
29:12
It's always the nature of these deals.
29:14
>> You got to you got to price in the the Gulfream expenses with the with the lobbying fee.
29:19
You can tell more about how a company is doing in DC by where their Gulf Stream is parked than their lobbying budget for sure.
29:28
Where's the Gulf Stream parked?
29:32
Has it been at Mara Lago? Has it been in DC?
29:33
That's what's moving the needle these days.
29:36
Uh this is this screenshot says Alman and Wong um uh negotiated their pact largely through a mix of virtual discussions and one-on-one meetings in London, San Francisco, and Washington DC with no bankers involved.
29:51
According to people close to the talks who declined to be named because they weren't >> This was a multinational deal.
29:57
>> It was they were going all over the world.
29:58
Uh London, they were probably talking about it in in White Tales.
30:03
Remember, they were in White Tai. uh and DC, who knows?
30:05
They might have been chatting to each other.
30:07
I don't think they were both at Hillen Valley, but they might have been there over overlapped at something or other, maybe the inauguration.
30:11
Uh the arrangement calls for Nvidia to invest 10 billion at a time in OpenAI. >> 10 billion slugs.
30:18
>> They have 60 billion in cash and cash equivalents right now.
30:20
So they can without generating any more free cash flow, they can just keep paying that.
30:26
>> Would if you were Jensen, would you would you lever up massively?
30:33
maybe >> become just entirely indexed to each other, right?
30:38
>> I mean, it's it's happening.
30:41
>> So, if we go down, we're going down together.
30:43
>> There's more there's more news about the the the AI kettu that's coming together.
30:48
OpenAI partners with Oracle and SoftBank for five new US data centers.
30:50
7 gigawatts is announced.
30:54
>> This and and what's crazy is they're still while they're doing this. Yeah.
30:58
>> They're doing deals with Broadcom.
31:00
They're doing deals with uh >> Oracle with coreweave >> Microsoft still I mean they're buying from everyone.
31:07
It's uh they don't want no one wants to be GPU poor in 2025 going into 2026.
31:11
You you want to be uh you know keeping the GPUs cool not on fire. Not on fire.
31:19
>> Uh so Wall Street Journal says OpenAI laid out its vision for a vast one trillion building.
31:22
I think it's better if if if you if if you're scaling GPUs but keeping them on fire if they're not on fire that partners are going to be looking and saying >> you don't want cold GPUs sitting there idle just eating a hole in your balance sheet.
31:36
You want to be maxing those tokens out >> getting paid.
31:38
Uh showcasing the development of a central park-sized complex 180 miles west of Dallas.
31:42
Uh this is from uh two days ago by Berber Jinn in the journal.
31:48
OpenAI unveils plans for seemingly limitless expansion of computing power in Texas prairie startup showcases ground zero of AI boom and its plans to shepherd one trillion in infrastructure spending.
32:01
So the trillion dollar number is getting thrown out thrown around for the first time because Stargate was originally 500 billion and now we're up at 1 trillion. Is that right? >> Yeah.
32:11
And remember even at 500 billion Elon was basically saying hey nobody involved here actually has the cash to do this.
32:18
So, >> just wait until these stocks run up and Oracle's up 300. Larry says hold my beer. I'm good for my 10. >> Hold my green juice.
32:28
>> I'm here for I'm I'm I'm good for my half trillion.
32:31
>> Um but yeah, once you get into the trillion dollar range, you're approaching the GDP of uh you know, some some serious countries, right?
32:40
>> Some serious countries.
32:40
OpenAI disclosed it would ultimately need more than 13 times the computer the computing power of its first nent site which is rais which is rising out of the Texas brushland.
32:51
Frenzied construction here has turned a sea of red dirt into eight hyper futuristic data centers bringing online roughly 900 megawatts of capacity.
33:00
More than 6,000 workers labor on the project each day, including electricians, plumbers, and steel welders alternating between two 10-hour shifts seven days a week. Wow.
33:09
They are they are uh what is that?
33:12
Uh 2 to 107 210 7 2 am to 10 p. m. 7 days a week something. >> Yeah.
33:21
People in the data center business see uh people with computer computer jobs saying oh yeah I do 996. >> Yeah.
33:27
I mean I wonder if they have two shifts 10 hours does that mean they have two crews that are working 70our weeks?
33:34
Are people taking days off or are they rotating their shifts so that you get a day off every >> I think they would get into pretty hot water.
33:40
I don't think you could like run and say you're working seven days a week forever.
33:44
I don't think I don't think employment law works like that.
33:47
>> Gray towers of gas turbines have dotted the landscape since the spring, offering backup power.
33:52
On a tour with reporters Tuesday, Oracle and OpenAI executives showcased the 1,100 acre site, calling it the largest AI supercomputing complex in the world.
34:01
Outside workers wearing dust masks and sunglasses shuttled around in buggies trying to shield themselves from the 100 degree heat.
34:07
One of the buggies flew a flag that said Jesus is the answer. >> Not very hgi pill.
34:14
I thought we were getting God in a box. What's going on?
34:21
Just Elazar Udicowski shows up. He's like, "No, no, no.
34:25
Like the building god here. You're building God. You're doing the thing.
34:29
Like what are you doing here?"
34:29
Uh there was literally nothing here a year ago uh said uh Anous who works on the OpenAI computing team.
34:37
They also announced five new data sites data center sites across the US built with Oracle and Japanese tech conglomerate SoftBank.
34:45
It said the new facilities would help bring online nearly 7 gawatts of power.
34:49
Enough for almost 8 million homes.
34:51
Enough for a lot of API calls from Devon, from Cognition, the makers of Devon, the AI software engineer.
35:00
crush your backlog with your personal AI engineering team.
35:05
>> Company executives made it clear that the Abene site was just the beginning, noting they envisioned a need for more than 20 gawatts of computing capacity to meet the explosive demand for chatt which now has more than 700 million weekly users.
35:16
Um, I was laughing about, you know, how in the chat GPT if you fire off pro, like GPT5 pro, it kind of thinks for like 20 minutes and everyone's saying like, oh yeah, this is good.
35:30
Like, we want the the agents to be reasoning for longer and longer.
35:32
I I feel like there's an opportunity for some startup just to be like, we've cracked the code.
35:36
Our model reasons for two days straight.
35:39
So, send us your request and we'll get back to you in two days.
35:45
>> And the agent just like is pretending.
35:47
So it's just like >> Yeah, it just has a progress bar >> twiddling it some until the last 20 minutes and then just starts going. >> No, no, no.
35:52
Of course, what you do is you is you have the loading bar there and then secretly you have a human doing the work and you have a human that's like dealing with this.
36:01
>> Well, I think the thing is like the longer the longer the the the period, the higher the expectations.
36:06
And I just don't know that I don't I I think that uh I guess the thing that I would want to know is like if you're running a deep research query for 20 minutes and then if you decide to run it for two days, how much better is it? Right. Yeah.
36:23
>> If it's >> if it's 10 times better, >> yep, >> could potentially be worth it for different types of tasks.
36:29
If it's >> 30% better, is it worth the wait?
36:31
Yeah, there there is a world where you just route it to the human.
36:36
There's also the world where for for the for the really low >> don't encourage this.
36:42
People have already gotten into trouble. >> Yeah, they have. They have.
36:44
Uh the the other funny idea is like for a lot of the there have to be a certain set of queries that just get asked the same like every single day.
36:55
And they should just store those in a database and cache them because if like you have to imagine that every single day someone is asking like just going to chatt and lighting the GPUs on fire to just ask like what's the capital of Illinois, you know, or what's the capital of California?
37:12
>> Can you run a deep can you run a a GPT like pro deep research report for what is the capital?
37:19
>> You can do it for anything.
37:19
Yeah, you can you can go to deep research and fire off 20 minutes of compute and say what is the capital of California answer in just one word and like it will just think for 20 minutes and then spit out Sacramento like >> no it won't think that long.
37:33
It it like thinks based off how hard it is.
37:36
>> No, deep research will >> I just said what is the capital of California answer in one word and it's >> deep research.
37:44
>> Yeah, it's doing a deep research query.
37:46
>> Let's see how long this takes.
37:46
I I I it's not going to it's not going to come back in in two minutes.
37:50
It's going to it's going to cook for 10 minutes. >> All right.
37:52
I was a little bit late, but I just started a a timer. >> Okay.
37:56
>> So, what do you think?
37:57
>> Also, there's like a sense where like um the like embeddings of facts is basically just the model already.
38:01
Like you can just think of a model as like a compressed version of the internet or at least like non-reasoning models. Yeah.
38:08
So, it's kind of like, >> but wouldn't it be even more compute efficient to just do fuzzy lookup of of these questions and see if they >> in the chat says cachebased retrievalss are >> Oh, yeah.
38:20
>> But apparently not here because >> GBD5 fully. >> Okay. Okay.
38:25
It reasoned for 47 seconds. >> Okay.
38:28
>> And it like >> this is this is pro.
38:29
You didn't do deep research though.
38:31
>> I want to see you turn on deep research and do it.
38:35
Well, this is the pro which is research grade intelligence. >> No, no, no.
38:38
That's not the deep research you thinking. >> No, no, no, no.
38:40
You need to click on the little plus button next to the chat box and then check the box on on uh on deep research. Do they still have this? I don't know.
38:51
They might have they might have deprecated this.
38:51
I think it's still there. >> It's still there. Still there. Okay.
38:53
I'm doing a deep >> You're doing research. >> No, no, this is good. This is good.
38:58
Verify >> just to confirm, are you asking for the current capital of California or or are you interested in historical capitals as well? >> Yeah, >> let's see. Current capital.
39:08
>> Tell it to also design a uh a website in Figma.
39:12
Think bigger, build faster.
39:12
Figma helps design and development teams build great products together. Get started for free.
39:18
Uh while we're waiting for that, let's keep reading from the journal.
39:21
Each gigawatt of capacity is expected to cost roughly 50 billion, Tyler. 50 billion per gigawatt.
39:26
That's the number in the journal, meaning the company is laying the groundwork for at least 1 trillion in infrastructure spending.
39:35
Demand is likely to reach closer to 100 gigawatts, one company executive said, which would be 5 trillion, >> which is roughly the GDP of Germany >> or Japan.
39:47
>> Yeah, GDP of Japan is 4 trillion. Germany is uh 4. 6. >> Good luck.
39:53
Good luck, Germany and Japan.
39:55
You're getting left behind.
39:55
You're part of the permanent underclass now. It's over for you.
39:59
>> I feel bad for lighting the GPUs on fire for this.
40:01
It says >> it's >> got it.
40:02
I'll confirm the current capital of California for you and get back shortly.
40:06
And it's >> it's doing it's cooking.
40:10
Like this is the real pitch for the uh >> I'm getting my money's worth.
40:15
>> Brandon Jacobe in the chat is breaking.
40:16
Jordy discovers the UX issues of every foundation model company because >> fix it Brandon. >> Yeah. Yeah.
40:22
Aren't you doing consulting now, Brandon?
40:24
You got to get in the in the trenches. Yeah.
40:25
All these folks, >> if the labs haven't hit you up yet, >> if someone who spends 10 hours a day studying this stuff, >> it's reading Wikipedia.
40:37
>> Just like how many drops of water did this use?
40:39
How how many fish had to like go find some other home?
40:44
>> It should come back and it's like thought for 15 minutes, put made two species extinct, >> lit lit seven gallons of diesel on fire.
40:55
Okay, it's checking all the sources. >> That's good.
40:58
I I like a detailed report.
41:00
I like to know that the AI did its job.
41:02
Don't just sit there clanker and riff off cashbased retrieval.
41:04
I want the deep research every single time.
41:09
>> Even for one word fact.
41:10
>> Even for one word fact, I want you to be sure.
41:13
I don't want any hallucinations in the >> Well, I mean, this is a this is a tough question, too.
41:16
A lot of people that aren't from the West Coast or the United States would say, "Yeah, what's the capital of California?" Yeah, San Francisco.
41:24
>> I was using this as an example about >> I remember being I must have been like seven years old asking like figuring out that Sacramento was the capital of California and I thought that was the funniest thing ever. >> Yeah, it is. It is hilarious. >> All right, it got it.
41:38
It thought for 101 seconds.
41:41
>> That's not that long for deep research.
41:43
>> It went for four sources. It did 19 searches.
41:45
So, this this is why this is why I can't trust, you know, everybody keeps screenshotting uh uh these Google trend reports. >> Yeah. Yeah.
41:55
>> I'm not sure that Google has figured out how to >> exclude >> exclude agents because this just did >> 19 searches for a single fact. Yeah. And so that's true.
42:07
>> If you're looking at the growth of queries on Google, you would think >> and and if a lot of people are asking this kind of question, every single >> keyword is going to be spiking.
42:19
>> Well, if you want to get your brand mentioned in chat GPT, you got to go to profound. That's right.
42:22
Reach millions of consumers who are using AI to discover new products and brands and get a demo.
42:29
Um, I don't think we've figured out the final form of what financing for compute looks like, said Sam Alman.
42:35
But I assume, like in many other technological revolutions, figuring out the right answer to that will unlock a huge amount of value delivered to society >> and delivered to OpenAI shareholders. >> For sure.
42:50
Yeah, it is it is an interesting question because you know Dario was talking about how like each training run looks great by itself but when you put them all in one structure it looks like you just have this neverending money pit because you spent a hundred million dollars on a training run.
43:06
You made a billion over the next two years.
43:08
You spent a billion on a training run.
43:10
You made 10 billion over the next two years.
43:12
You spent a hundred billion on a training run.
43:13
You make a trillion over the next two.
43:15
And there's a question, okay, how long until it just completely maxes out?
43:18
Are there diminishing returns to these things?
43:20
But even if the trends hold, you still wind up with this like ever growing money pit that should act as like some sort of slowing down force.
43:30
You would >> I think when you when you dig into this quote, I don't think we figured out the final form of what financing for compute looks like.
43:36
The question is like will this actually be a novel financial instrument or it will be remixing something that Wall Street has used in mortgages? >> Maybe it's debt. Maybe it's debt.
43:47
who knows >> GPU back securities.
43:51
>> Uh the Tuesday announcement made it clear that SoftBank once seen as a formidable OpenAI funding partner has scaled back its ambitions in the data center buildout which the Wall Street Journal reported in July.
43:59
Three new sites, one located near Abalene, another north of El Paso in New Mexico and a yet to be announced Midwest location combined with an expansion to the Abene complex will be capable of delivering 5. 5 gawatts of capacity.
44:11
Those will be built by uh Oracle.
44:14
The other two smaller sites, one in Lordstown, Ohio, and the other near Austin, Texas, will be built in partnership with SoftBank and generate 1.
44:22
5 gawatts over the next 18 months.
44:25
The first completed data center called building one painted a pristine white that contrasts with the reddish dirt surrounding the site is larger than two Walmart Super Centers.
44:35
Doesn't feel that big to me, but certainly huge. Uh entering.
44:38
Yeah, the the 1100 they said, 1100 acres earlier in the article.
44:44
That's under two square miles. >> It is.
44:47
Yeah, it is crazy how compressed this is.
44:49
Obviously huge when you're walking around, >> but you know, we're not in Okay, we're we're we're losing yuseite over this yet.
44:56
>> Yeah, we're not in the, you know, >> in the full terraforming.
44:58
I mean, I said told the what was it the San Francisco Chronicle that we could he could imagine a future where the entire earth is covered with solar panels if the trends continue.
45:09
Gab Gabe in the chat says, "Shout out to Lordstown, Ohio.
45:12
Factory used to build Chevys."
45:15
This is your your thesis or or take re-industrialization is happening.
45:21
Just it's a lot less jobs. Yep. And it's data centers. >> Yep.
45:26
Um >> but the the the uh white pill there might be that uh if we can build big data centers really quickly in in uh you know at at the scale that traditional infrastructure projects would have taken 10 plus years.
45:42
If we can do that in a few years >> it could create a precedent where somebody could decide you know what I'm going to build this this new rail line in two years.
45:50
Is a is a Walmart super center really one mile long?
45:53
Uh QBS rollover is saying that um that the uh that the building's 2 miles long that I mean >> well no I was saying I was saying the site is is 1100 acres total.
46:05
It's not uh >> uh which is like 1 1. 7 uh square miles.
46:13
So obviously massive but I don't think the building itself is uh anywhere near that.
46:18
Well, fiber lines snake across the data centers and underneath the ground, allowing the AI chips called GPUs, in case you were wondering.
46:25
The Wall Street Journal's got you covered uh to talk to one another and complete requests more quickly.
46:31
Proponents of the infrastructure boom say it will bring hund hundreds of thousands of jobs and revive American manufacturing.
46:36
In January, OpenAI unveiled a 500 billion dollar data center uh data center project.
46:40
Stargate, we know about this.
46:42
Uh the reality is more mixed.
46:43
While data center providers uh provide plentiful temporary construction jobs, far fewer people are needed once they once they are built.
46:50
Abalene mayor uh said residents had mixed feelings about the site and its power and water usage, though some of the concerns have been assuaged.
46:56
An Oracle executive said there will be roughly 1,700 permanent jobs on site once the construction ends.
47:01
once the construction ends. I think the things that ends up being a little bit yeah disappointing is is these projects get announced and they're saying we're going to spend $20 billion here and then we're going to create like
47:15
a thousand jobs or we're going to create 300 jobs and that is very significant that will be good for the local economies uh in these regions but not it's not hey we're yeah >> we're repl you know you we we lost 87,000 manufacturing ing jobs last month, I think it was. >> We're not really making a dent in that
47:34
>> We're not really making a dent in that data center. >> It takes a lot.
47:37
Um, there's more news on the job displacement stuff from Andre Carpathy.
47:43
He has a good post breaking things down.
47:46
But first, let me tell you about graphite. dev.
47:47
Code review for the age of AI.
47:49
Graphite helps teams on GitHub ship higher quality software faster. Get start for free.
47:52
So, Andre >> Carpath debating by the way.
47:56
>> What is Chad debating?
47:57
>> They said uh Jord is a Cadillac. >> I'm a Cadillac. True.
47:59
John has >> Somebody said, "I I seem like a Dodge guy." >> Hellcat SRT. I love it.
48:07
>> Uh yeah, not not quite. Never had a Dodge.
48:09
I had a I had a Ford Raptor, which I loved.
48:11
Uh was not super practical.
48:14
Uh but uh it was a car that I wanted as a child, so I had to get it at least once. >> Yep.
48:20
Um so Andre Carpathy says, "AI isn't replacing radiologists."
48:27
Uh Tyler, do you remember this prediction? >> Yeah. Was this Hinton? >> Hinton? Yeah. Yes. Uh years ago.
48:31
This was this was what 10 years ago now or something. >> Yeah.
48:37
And because this was based off just like image classification.
48:39
So this was like early on like CNN's like this was probably mid2010s I assume.
48:43
So uh the expectation was that rapid progress in image recognition AI would delete radiology jobs as famously predicted by Jeff Hinton now almost a decade ago.
48:54
Uh in reality radiology is doing great and is growing. So what's going on here?
49:01
Obviously image recognition is fantastic.
49:03
These algorithms are remarkable.
49:05
Um and there's been a ton of progress.
49:07
So Andre Carpathy says there are a lot of in his opinion naive predictions out there on the imminent impact of AI on the job market.
49:14
For example, a year ago, uh Andre was asked by someone who should know better if he thinks that there will be any software engineers still today, like a year from now.
49:27
Like this is how accelerationist this person was.
49:29
Uh and Andre says, "Spoiler, I think we're gonna make it. This is happening.
49:34
This is happening too broadly."
49:36
Um, this post goes into detail on why it's not that simple using the example of radiology.
49:41
And I thought this was really really interesting.
49:42
Um, and it obviously applies outside of radiology.
49:45
So, first off, the the benchmarks are nowhere near broad enough to reflect real actual scenarios.
49:51
So even though you can classify cancer in a radiology environment with a whole bunch of training data very reliably, there's still tons of edge cases where having a human in the loop is advantageous.
50:04
Uh and then then and then just in general the job is a lot more ma multifaceted than image recognition.
50:12
Like if you just think about the job of a radiologist, it's not purely just look at image detect cancer, look at image, detect cancer, look at image. >> Yeah.
50:23
Can you can you imagine if if you were getting uh god god forbid somebody's getting uh seeing going and getting uh seeing a traditional office of a radiologist and and they're just watching a television screen that's reporting telling them that something is wrong with them using a voice model to explain it and then it just is like >> no they're just in hinge mode.
50:44
Swipe right if it's if it's cancer, swipe left if it's not.
50:48
And they're just sitting there swiping all day long.
50:49
Obviously that's not the job of radiologist. There's so much more.
50:53
They're doing research on what's changing, integrating new information, talking to c talking to different uh patients about all the different signals, what they could be doing, what they could be putting in their body, how are they feeling, uh what medicines they're taking, there's a ton of different stuff that they're probably involved in. >> Yeah.
51:10
Anytime anyone gets a any type of lab or report from a doctor, >> y >> the first thing you want to do is talk to somebody that's an expert on that, your doctor, and have them explain it to you, the implications, etc. , good or bad.
51:23
>> And then he highlights deployment realities.
51:25
So regulatory, insurance, liability, diffusion, and an institutional inertia like it just takes time to roll these technologies out.
51:32
But the last one is super interesting to me.
51:36
He cites Jeban's paradox.
51:36
So if radiologists are sped up via AI as a tool, a lot more demand shows up.
51:41
And so if all of a sudden the the cost to get a scan or the time that it takes to actually do the scan drops, everyone will say, "Well, I want to be scanned all the time.
51:55
I I I want to be on top of uh any sort of development medical that could be detected by a radiologist."
52:00
And so, uh, he concludes by saying, "I will say that radiology was, in my opinion, not among the best examples to pick on in 2016.
52:09
It's too multifaceted, too high risk, and too regulated.
52:13
When looking for jobs that will change a lot due to AI on shorter time scales, I'd look in other places, jobs that look like repetition of one wrote task, each task being relatively independent, closed, not requiring too much context, short in time, forgiving, the cost of a mistake is low, and of course, automatable given current and digital capability.
52:36
Even then, I'd expect to see AI adopted as a tool first where jobs change and refactor, more monitoring and supervising than manual doing, etc.
52:48
Maybe coming up, we'll find a better and broader set of examples of how all of this is playing out across the industry.
52:54
About six months ago, I was asked I was also asked to vote if we have if we will have less or more software engineers in five years.
53:02
Exercise left for the reader.
53:04
What do you think he meant, Tyler?
53:07
What's what do you think >> more of course he he thinks the number of uh of software engineers will increase but they will be using AI as a as a tool to leverage. >> Yeah.
53:19
>> Yeah. Yeah, I wonder if anybody's done a study on on how many more developer people that that uh have done software development this year versus last year has to be significantly more just because n you know if you're using replet now right you are
53:36
>> a software engineer right it's it's the >> somebody might want to get technical and say oh well if you're just prompting you're not really software engineering but how different is that from somebody getting their start in like a copying pasting, you know, lines of code. >> Wow. Uh we we got some news about Tik >> Wow.
53:51
Uh we we got some news about Tik Tok that uh China has like more or less signed off on the deal and it has sent Larry Ellison/Oracle skyrocketing on Poly Market to an 89% chance of acquiring Tik Tok.
54:05
Uh this will be an interesting story to see where it actually lands.
54:10
Um, of course you can go to Poly Market if you want to keep monitoring the situation as always.
54:18
Um, >> should we talk about this uh new Elon lawsuit?
54:22
>> Yes, new Elon lawsuit.
54:24
>> Elon Musk is uh Andrew Curran uh is on the timeline sharing Elon Musk is suing OpenAI again.
54:32
second suit uh this time for alleged misappropriation of trade secrets, intentional interference with pro, uh prospective economic relations, and unfair competition.
54:42
Um so they say the desire to win the artificial intelligence race has driven OpenAI to cross the line of fair play.
54:49
Opening eye violated California federal law by inducing former XAI employees including Shu Shen Lee and Jimmy Freighter and a senior finance executive to steal and share XAI's trade secrets.
55:02
Uh the the copy in here is funny.
55:05
Uh and and again this is like from the this is from the complaint, right?
55:09
The lawyers are >> not using chat GBT for this is uh this is hand handmade by hook or by crook.
55:18
Open AI clearly will do anything when threatened by a better innovator including p plundering and misappropriating the technical advancements, source code and business plans of XAI.
55:27
What began with OpenAI's suspicious hiring of Shu Chen Lee, an early XAI engineer who admitted to stealing the company's entire codebase.
55:36
>> Wow, >> that's bold thing to admit, has now revealed a broader and deeply troubling pattern of trade secret misappropriation, unfair competition, and intentional interference with economic relationships by OpenAI.
55:48
OpenAI's conduct in response to being out innovated by XAI whose Grock model overtook OpenAI's chat GPD models and performance metrics reflects not an isolated lapse but a strategic campaign to undermine XAI and gain unlawful advantage in the race to build the best artificial intelligence models. What do you think John?
56:07
>> This is why lawyers with English degrees are MVPs according to Gabe. >> Yeah. Yep. Wordsmiths, word cells.
56:11
I it seems like uh we are in the territory of like economic warfare, lawfare uh who knows what this what actually comes >> blending the the Chinese way the way of the engineer with the American way. >> Yes.
56:28
The lawyerly society is coming out in this case.
56:30
Um >> yeah I mean easy you know obviously in innocent until proven guilty with Shu Chen Lee.
56:37
uh don't know if this is true or not that he admitted to stealing the company's entire codebase would be crazy and totally unforced error.
56:45
Uh if he actually did something like that. >> Yep.
56:51
>> But um >> well no matter what you want to stay compliant, you want to get on Vant automate compliance, manage risk, improve trust continuously.
56:56
Vanta's trust management platform takes the manual work of the dirty compliance process and replaces with continuous automation.
57:03
whether you're pursuing a first framework or managing a complex program.
57:06
Um, I wanted to revisit this internal tech email because this is uh is an important cornerstone of Elon's battle with open AI.
57:17
So, uh, on September 20th of 2017, Elon Musk sent an email to the OpenAI team.
57:23
Uh, the subject was honest thoughts and he said, "Guys, I've had enough. This is the final straw.
57:31
either go do something on your own or continue with OpenAI as a nonprofit.
57:36
I will no longer fund OpenAI until you have made a firm commitment to stay or I'm just being a fool who is essentially providing free funding for you to create a startup. Discussions are over. Interesting.
57:47
And it's gone back and forth because here he's he's sort of predicting.
57:53
So in hindsight, he somewhat probably feels like a fool for essentially providing free funding for you to create a startup.
58:03
>> So, >> and it's it's been odd because the the ask has never I mean, at least that I've been aware of.
58:08
Elon's never just asked for proada equity based on how much he put into the nonprofit.
58:13
I think that creates crazy tax and legal implications.
58:17
But that feels like what the fair ask would be.
58:18
the fair ask would be. you like you were throwing out like uh if you put in a hundred million and you just assume that the the round that would have been done at that time would have been a billion you should get 10% right and here's 10% of the common or the preferred or whatever
58:34
>> remember nobody like it seems like they hit a standstill with the forprofit conversion right >> because of Elon's lawsuit too >> well part partially but also people are you know the the uh governing body the nonprofit governing bodies of California were not super excited about the transition. >> Yeah. And so Matthew Berman um was going >> Yeah.
58:53
And so Matthew Berman um was going back and forth with Elon uh because Elon took shots at Anthropic.
58:58
He said winning was never in the set of possible outcomes for Anthropic.
59:02
And Matthew Burman said, "Hey, Elon said the same thing about OpenAI."
59:06
And Elon said, "No, I never said that.
59:08
I told everyone who asked in the beginning that I thought the probability of OpenAI beating Google was 1% infinitely different from 0%." Good point.
59:18
And uh and then Elon says after the Defense of the Ancients, the Dota 2 win, I raised the probability to 10%.
59:26
That they would catch up.
59:26
And Matthew Burman sort of responded and unpacked the screenshots saying, "First, I'm a big fan of everything Elon has done to push society into the future, but I reread his emails with Sam Alman and Greg Brockman, and it really seems like he told them they had no chance of succeeding relative to Google on January 31st, 2018."
59:45
And so after that previous email we read, Elon told uh Elon told Sam Alman and Greg Brockman, OpenAI is on a path of certain failure relative to Google, there are obviously needs there obviously needs to be an immediate and dramatic action or everyone except for Google will be cosigned to irrelevance.
1:00:07
And it's played out very differently.
1:00:09
They played their hand very well.
1:00:09
They got the consumer application that hit product market fit out faster than Google only by a couple months.
1:00:16
Like the Gemini app dropped I think in February or March just a few months after Chachi PT launched and the models have been neck and neck.
1:00:28
It wasn't I don't was it Bard that dropped? I forget which.
1:00:32
thought Bard was the response.
1:00:33
>> But I I feel like in the the way I remember the timeline is that the chat GPT website comes out and everyone realizes that just talking to an RHF LLM is a great experience and magical and interesting and passes the touring test.
1:00:49
And Google had a bunch of models palm and barred internally, but they'd never really pushed them out in a meaningful way.
1:00:55
And it seemed like Chachi PT kicked Google into action and led Google to actually like productize a lot of their research that had been more tested and shared internally.
1:01:07
Tyler, do you have more context there? >> Uh, yeah.
1:01:09
So, it was Bard that came out right after Chedd and uh, Gemini was like the following. >> Yeah.
1:01:15
So, they still had to do a rebrand which is tough because then you have to reintroduce the product.
1:01:18
Obviously, it's it's grown a ton and we've seen that, you know, the empire strikes back.
1:01:22
Not a lot of that happening in the United States necessarily, but certainly international.
1:01:27
Sort of an iPhone Android story happening all over again.
1:01:29
Um, anyway, uh, Magnus Carlson tells Joe Rogan about how he treats chess as a hobby.
1:01:37
He never wants it to feel like work and prioritizes fun and enjoyment.
1:01:41
The thing is that chess has always been a bit of a hobby for me.
1:01:44
He's one of the best chess players in the world, of course.
1:01:47
Uh, this is how Magnus Carlson works.
1:01:50
Once it starts to feel like work, then it's harder for me.
1:01:53
Turn your work into fun and enjoyment.
1:01:55
You'll never work a day in your life.
1:01:59
>> You might get 14,000 likes on X, like like Dylan. >> Yes.
1:02:02
And uh and Daniel growing Daniel says, "Having known him for a while, I think Elon Musk is probably one of the least self-disciplined people on Earth.
1:02:12
He just really enjoys problem solving.
1:02:14
So, he just bounces around from one project to the next solving it. Moging them." Yes.
1:02:19
Uh, of course, uh, uh, Growing Daniel's co-conspirator now runs Julius. ai.
1:02:27
What analysis do you want to run?
1:02:27
Chat with your data and get expert level insights in seconds.
1:02:30
Loved by over two million users and trusted by individuals at Princeton, BCG, and Zapier.
1:02:34
Uh, Bas Lord says, "Anyone who's naively calling bubble right now has not internalized the physics of our world.
1:02:42
Most of them are simply ignorant about AI progress and its implications.
1:02:44
Of course, there will be fluctuations, but of course, we could stumble.
1:02:48
But history marches on and he shares a chart of the world GDP over the last two millennia.
1:02:53
And it's interesting because like you can take this in a few different ways.
1:02:59
You can you can kind of look at this and say, okay, nothing ever happens.
1:03:02
Um it's just a kind of a smooth exponential curve.
1:03:08
But I'm looking and I'm seeing a hiccup there in like the mortgage back security crisis for sure where GDP shrunk globally or something happened.
1:03:17
>> But it's it's helpful. Well, it's helpful. >> Isn't that 2007208? Right.
1:03:21
>> You know, you remember being uh I was, you know, >> basically a kid at that time and it felt like I could sense the the doom and dread in the in the adults right around that time.
1:03:35
>> I mean, from my perspective, I was completely caught off guard by the by the mortgage back security crisis.
1:03:38
Like, it was because I wasn't following the market. You didn't have capital.
1:03:42
I wasn't monitoring the situation. Yeah.
1:03:45
>> You didn't have capital to deploy by.
1:03:45
I lived in a house with a mortgage.
1:03:46
So I was certainly affected but um and uh but but back in time you would have you would have you know raised and deployed a fund into single family >> but I wasn't I wasn't seeing top signals.
1:04:00
It was like I was introduced >> where were you in college?
1:04:04
>> I had I it the crisis I graduated in 20 in 2007.
1:04:07
So uh >> from college >> from high school no from high school.
1:04:12
And so the when I arrived I chose to study economics because I wanted to understand the crisis more.
1:04:17
And so I was reading the Wall Street Journal trying to understand and I was introduced to like the collapse of Lehman Brothers which I believe happened in 2008.
1:04:23
Uh uh Lehman Leman Brothers collapse. When did that happen? The bankruptcy. September of 2008.
1:04:32
And so I graduated in 2007.
1:04:32
And then so it must have been like sophomore year.
1:04:39
So things were like melting down my freshman year, but it was kind of >> collapse of Lehman. Is that good? >> Is that good?
1:04:46
>> Did we not like Lehman?
1:04:47
>> I mean, it was it was a very complicated crisis to dig into because it was all these layered financial instruments, mortgage back securities, and then collateralized debt obligations.
1:04:55
So they would take one mortgage, they take a whole bunch of mortgages.
1:04:59
This is the famous like Margot Robbie in the bathtub scene where she explains that uh you get a whole bunch of mortgages, you bundle them up, you securize, you slice the different tranches of debt so that the best mortgages are in the AAA and then but then as you keep re-slicing them, they kept getting rerated as oh well this is the best of the previous bad stuff, so that's good.
1:05:22
And the best of the worst is the best.
1:05:24
and uh and the the rating agencies had a had a really bad time throughout the uh because eventually like the economy pulled back.
1:05:34
Everyone got over levered.
1:05:34
people were buying multiple homes and there was a whole mood in 2005, 2006, 2007 that um like average people like the the example would be you know when the when the the Uber driver is pitching you on Cardono like it might be the top that type of thing like when when when you know your barber is telling you about the latest NFT drop like watch out.
1:05:56
Uh the the the analogy then was people were saying like I'm making more money just owning my home and my home equity is going up by more than my salary.
1:06:07
So like I bought a $500,000 home.
1:06:09
I make $100,000 a year, but my home my my banker just called me and said, "Your home's worth $600,000 now."
1:06:17
And the home prices were going crazy.
1:06:18
And so >> obviously there had to be a correction and it crashed the entire global economy.
1:06:24
And you can see it in the data today.
1:06:26
But of course as we zoom out further uh we will you know see less >> line keeps going up.
1:06:33
>> The line keeps going up.
1:06:33
Well um DTOC is cyclical but ever evolving.
1:06:36
The back and forth locked in shares this chart or this uh meme.
1:06:41
Drop ship bros convert to supplement bros which switch to drop ship bros which convert to supplement bros which go back to drop ship bros and back to supplement bros.
1:06:50
Um, >> supplements continue to be the economic engine of the manosphere.
1:06:57
>> Weren't we talking to the the CEO of uh of Numel and Sam and he was saying he owns a supplement company.
1:07:05
>> Yeah, he he started a gummy supplement company back in the day that's still he's like it's still doing well figures. That's crazy.
1:07:10
Anyway, numeral sales tax on autopilot.
1:07:12
Spend less than 5 minutes per month on sales tax compliance.
1:07:14
If you're a drop ship bro, if you're a supplement bro, you got to pay your taxes.
1:07:19
You got to get on numeral. Great. Great.
1:07:22
Dan McCormick's company is on numeral.
1:07:24
>> Yeah, >> he's a supplement bro. >> Lucy is on numeral.
1:07:26
Um >> Christina Cordova is quoting uh our interview with uh Christinez over at Linear, of course.
1:07:34
Uh but we asked Brett Taylor yesterday about the triple triple double double debate, uh and how maybe that's not good enough anymore.
1:07:44
Uh he said, "The faster you grow is sometimes correlated with lack with a lack of a moat.
1:07:50
I'd rather bet on a durable high quality revenue uh business and happy customers than just raw growth.
1:07:55
Uh Christina says fast ARR can be fragile ARR.
1:08:00
Earned ARR tends to be sticky ARR.
1:08:04
Not all growth is created equally.
1:08:04
I think linear is uh obviously they're a partner.
1:08:09
So I'm somewhat biased but uh uh but of course they have >> just for those who don't know linear is a purpose-built tool for planning and building products.
1:08:16
meet the system for modern software development, streamline issues, projects and product roadmaps but continue >> use the tool love by open AAI Mercury retool ramp >> scale anymore.
1:08:26
Um but uh linear is a perfect example of like great growth but steady and just systematically earning the trust of pretty much every important new company in Silicon Valley over the last however many years.
1:08:42
I think they're um >> yeah earned ARR.
1:08:44
Yeah, it does feel like there's a 2019 there's like C ARR reared its head recently.
1:08:51
It's kind of gone out of favor, but uh contracted ARR basically you have some sort of LOI or some sort of trial that you're counting into >> Yeah.
1:09:02
I mean the the the the people that are abusing the ARR the most right now >> extremely disrespectful to the gap metrics.
1:09:12
Well, well, and I'll just say it's it's the uh companies doing data labeling and services for labs, right? There's >> is that ARR?
1:09:21
>> All the companies that exploded out of uh >> I don't think that many of them are actually quoting ARR numbers though, >> are they? >> Are you kidding?
1:09:28
I feel like almost every single one of them is saying, "Yeah, we got to 100 million AR. We did this or that."
1:09:33
>> I thought it was just like if you did 100 million because you just say, "I did 100 million, then that's true." >> No.
1:09:38
>> No. and they're saying we're at 100 million ARR and and they're and they're using it I think as like annualized run rate but if you dig into the business it's like they might have 40% margins because they're actually having to hire
1:09:52
>> well the margin separate from like AR with bad margins completely separate from ARR with bad churn >> like yeah because I would I probably if you're underwriting it at a high multiple you probably want to pay for lower margin lower churn than higher higher margin, higher churn. Because
1:10:08
Because higher churn means that like if the business stops growing, it's going to start shrinking.
1:10:14
Whereas if you have if you have 40% margins but super low churn, you're super sticky.
1:10:18
It's like, okay, well then at least the business will continue to grow.
1:10:22
You're not building on a leaky bucket, you know, you're not you're not getting all the juice out of the bucket.
1:10:27
You're only getting 40% of the juice since we're mixing analogies now.
1:10:33
>> But >> got the blender out these analogies.
1:10:34
Um but I mean obviously these two these two metrics are linked and you obviously want uh you know high margin and extremely steady revenue.
1:10:44
>> But if you're a company like linear and you see some which which has like true AR earned ARR Yeah.
1:10:51
>> and then you have uh companies out there saying you know throwing out these numbers around ARR but they're reselling uh re reselling GPU credits or something like that.
1:11:02
It's it's kind of, you know, stolen valor, right?
1:11:03
From the potentially true enterprise, you know, contracted uh >> uh >> Yeah. Yeah.
1:11:10
where you're you're you're like like you could, you know, stop the growth strategies, stop the sales team, they could stay home, and you're and you're still printing. Yeah.
1:11:22
>> That's the that's the that's the gold standard of enterprise software.
1:11:23
That's why we love >> that's why we get up every morning.
1:11:27
>> Uh we got to pull pull up this video uh from Paula Rambles.
1:11:30
There's this guy on Tik Tok who video calls ChachiBT shows objects asks how much Graham then weighs it.
1:11:38
Uh before we pull up this video, let me tell you about the generative video platform for developers.
1:11:43
Fall >> the world's best generative image, video, and audio models all in one place.
1:11:48
Develop and fine-tune models with serverless GPUs and on demand cluster.
1:11:51
>> You can get access to our soundboard by going to tbpn.
1:11:53
com/sounds, which is brought to you by our friends.
1:11:57
>> So, let's play the video of this guy on Tik Tok. This is my hand. How you call this thing?
1:12:04
>> Yeah, I see your hand there holding on to that pen. So, >> okay.
1:12:08
>> How much gram is the pen? How much gram? Look, how much gram?
1:12:14
>> Well, it's a little tricky to guess the exact weight. >> Guess guess.
1:12:18
>> All right, I'll take a guess.
1:12:18
For a pen that size, it's probably somewhere around maybe 500 to 700 g, give or take.
1:12:26
>> See if GPT5 is smarter. No, no good. No good. >> Hey, it's all good.
1:12:32
Honestly, >> it's no good. It's no good, man. It's no good. >> No worries at all.
1:12:38
It's just a bit of fun guessing. >> Don't worry. I'm worry.
1:12:40
Look, I have this oil here. How much g?
1:12:45
>> We're on a guessing roll.
1:12:45
So, for that little oil can, I'd say it might be somewhere around 300 to 500 g depending on how full it is.
1:12:52
But let's see how close we get this time. >> You said 300 to 500.
1:12:56
Man, I want accurate thing. You know what I mean? >> I get it.
1:13:02
You're looking for something spot on.
1:13:03
It's a bit of a challenge without a scale on my side, but hopefully that at least gives you a decent range.
1:13:10
>> This toana, how much gram?
1:13:14
>> Oh, you've got a mocha pot there. Nice.
1:13:17
Those little >> How do you call it? How you call it?
1:13:19
>> That's called a mocha pot.
1:13:19
It's basically a >> mocha pot. How much gram mocha pot?
1:13:26
So, a mocha pot is usually pretty light.
1:13:28
I'd say maybe around 400 to 600 g depending on the size.
1:13:30
But of course, you'll know. >> I hope you're right. >> What's the number? >> I hope you're right. >> Ah, >> bro.
1:13:38
Like, GPD5 is is worse and worer than GPD4.
1:13:42
>> Definitely here to do my best, and I appreciate you rolling with it.
1:13:45
>> Hey, this is your punishment, man. Talk with him.
1:13:48
>> Hey, Green Coco, any idea where the mayo's hiding?
1:13:50
Maybe it's just chilling out behind you in the fridge somewhere.
1:13:53
What's the deal with the ketchup in the mail?
1:13:54
Any condiment gossip we should know about? >> They just talked. >> That's hilarious.
1:14:00
That is not a Roco Bassilis respectctor. You should read that. That's very rude.
1:14:06
>> That guy's absolutely cooked.
1:14:07
>> Did Did anyone see exactly what the numbers were? How close was he said? Three. >> No. 30.
1:14:12
I I was seeing 35 on the >> That can't be 35 grams, right? It needs to be. I have no idea.
1:14:18
I'm worse than GPT5 for sure.
1:14:20
If you ask me how much that pan weighs, I have no idea.
1:14:26
Did anyone have any idea? I have no idea. Um, >> who knows?
1:14:30
>> Anyway, HBO Max says, "Nice.
1:14:30
We still got a couple of months."
1:14:34
>> Scroll down on this chart so people can see it's charting 1998 to 2001. >> Like, what?
1:14:39
Wait, this is just the NASDAQ and it's just following perfectly.
1:14:42
Like, what are we doing here?
1:14:45
This is crazy how how the correlation here is like insane.
1:14:49
Is this on the same like time scale too or is this this is just a wild wild chart that this so similar?
1:14:58
>> Um it lines up scarily accurately.
1:15:02
>> Um but it's just so weird because like if if it's this obvious you' think it would be priced in.
1:15:06
You think people would pull back or do something but who knows?
1:15:10
We'll keep monitoring the situation.
1:15:12
Um it will be interesting.
1:15:15
Let me tell you about turbo puffer. We puffing.
1:15:17
We search every bite serverless vector and full text search from burst principles on object storage.
1:15:22
Fast 10x cheaper and extremely scalable.
1:15:24
Jordy loves his puffer fish.
1:15:28
>> This puffer is used by linear notion many more >> WPP.
1:15:32
Wait, one thing one thing on this chart to be clear.
1:15:35
>> I I I think uh it it's tracking very closely.
1:15:39
Obviously, you you single out individual, you know, kind of uh >> points on the chart and it's a little bit ominous, but at the same time, like Amazon went public in 1998, right?
1:15:49
So like you can do a lot to kind of like if if you're the the the chartsmith here to kind of like pick the dates to try to align it properly, right?
1:16:01
So >> yeah, >> um >> I would I I bet you could align this to just other bubbles throughout history and find similar patterns, right?
1:16:11
>> That doesn't really help though.
1:16:11
I mean that that that's >> I'm just saying like a couple months. >> Yeah.
1:16:17
I mean I'm just I'm just saying like >> No, no, it it does feel like it's like if you started in 199 okay that the dot bubble began when Amazon went public right >> then this doesn't line up at all. Right. >> Sure. Sure. Sure. Yeah.
1:16:34
you're you're you're kind of just selecting.
1:16:35
I mean, yeah, I don't know.
1:16:38
It's Yeah, they're like overlaying it deliberately as close as possible, but this does feel like the crazy part is that the the the previous earlier selloffs, the the dip in 2023 uh to now is like it that pe that valley that peak and valley seems to match as well.
1:16:56
And then even like you can see earlier in 2025 we got the tariff selloff and that kind of tracks what happened in 200 uh in like the year 2000 exactly or maybe 99.
1:17:06
Uh lots of people >> yeah but then again was there like I don't know I don't know something you should read.
1:17:17
>> Even if we wind up in disillusionment we'll be slopping it up in the trough of disillusionment grinding to the plateau of productivity. one.
1:17:26
>> We'll probably end up in a position where like we're only going to eat fast casual slop bowls until we until we >> When you're in the trough, you eat the slob >> until we get out of the bare market. >> Eat the sllo bowl.
1:17:35
Uh >> somebody somebody had a great I got to I got to pull up this um >> comment on uh Spotify.
1:17:43
>> Someone's asking how how much the the gong weighs. Gong, how much Graham? >> How much Graham?
1:17:48
And Aqua says, "Mania is a crazy."
1:17:51
>> So, uh, Ma Flouse on Spotify says, "Safe out there.
1:17:56
>> Wearing the white suits on an off day to rally the bull.
1:17:58
Ultimate top signal might just be poking the bear.
1:18:05
>> Well, I'm in a gray suit today.
1:18:05
I'm feeling neutral about the market.
1:18:07
Could could go up, could go down. Who knows?
1:18:11
But we'll be monitoring it here."
1:18:11
Uh, in other news, breaking from us, TBPN, WPP media has signed Michael Mirllor as global EVP.
1:18:20
So, congratulations to Michael Mirlo.
1:18:24
He has been a fantastic champion. >> Massive pickup.
1:18:27
Hopefully wore his TVPN hat to work today.
1:18:31
Hopefully, he's not in the chat right now.
1:18:33
First day first day of work gets a corner office.
1:18:37
>> Puts TBPN on the first fired up.
1:18:37
Leave it on in the background. put titles on.
1:18:43
>> Uh but uh big pickup um and excited to see what he what he cooks on.
1:18:48
>> Always enjoyed his commentary and excited to see more of it.
1:18:51
>> Paul Enright says in 2021, Mitchell Green from Lead Edge Capital.
1:18:54
Of course, Mitchell is a friend of the show.
1:18:55
We got to have him back on again.
1:18:58
>> I texted him this morning. Um get him on soon.
1:19:00
Uh he said uh said to Paul, "This year will be remembered for the GPS that sold stock and distributed capital to LPs versus the ones that raised funds, called capital, and made new investments."
1:19:11
Uh was a wonderful observation I will never forget.
1:19:14
Now feels similar in some ways. Why not both?
1:19:18
Why not distribute capital, sell some positions, and raise some new funds?
1:19:24
>> Yeah, I don't know the dynamics of that, but that does feel like the the ultimate move is raise a massive fund at the top. Don't deploy it.
1:19:32
Sell a bunch of positions.
1:19:34
Get out of the other stuff.
1:19:36
Basically, just be sitting on a massive cash pile when you reach the trough of disillusionment and then monetize the the plateau of productivity.
1:19:44
>> Deploy into the trough. >> Into the trough.
1:19:46
>> Well, you saw this chart yesterday.
1:19:46
Um, >> uh, somehow like one of Andre's slides from a fundraising death. >> Newcomer, baby. >> He leaked master. Oh, yeah. He got two decks.
1:19:56
So um uh A16Z returned 15 billion or or or generated 15 billion in returns including their recycled fees and carry >> in 2021. Yep.
1:20:13
>> Uh that was you know biggest year on record by uh uh quite a lot >> huge return.
1:20:18
>> Um >> uh that's got to be a lot of Coinbase. And who else got out?
1:20:22
>> I think it was like Salana, Avalanche, right?
1:20:24
Some of these big >> a lot of stuff. >> I don't know.
1:20:26
Yeah, >> but Coinbase was a huge position >> and I mean they deployed in every all these different rounds and then uh I believe Coinbase went out and was probably distributed at something close to hundred billion market cap.
1:20:37
So um something something there.
1:20:41
Um Andrew Curran has more updates on Elon Musk's reuniting with the Trump administration.
1:20:50
Andrew says their friendship is slowly being reforged.
1:20:52
OpenAI, Anthropic, and Google already have similar agreements in place.
1:20:56
Uh, OpenAI and Anthropic charge a $1 access fee.
1:20:59
Google charges 47 cents because it's president is the 47th president, right? Isn't that it? That's interesting. Uh, that's cute.
1:21:08
Uh, Elon is charging the government 42 cents for access to Grock.
1:21:14
This is a Hitchhiker's Guide reference.
1:21:15
Have you >> finally a competition for who can say the smallest number >> and a lot of who can say the biggest number.
1:21:22
You want to be on the P on the barbell, the extreme ends, charging the least and raising the most.
1:21:28
>> Apparently, that's how business works. >> Most capex least.
1:21:31
>> Have you seen The Hitchhiker's Guide to the Galaxy? >> Have you? >> No.
1:21:37
>> I thought it was just a book.
1:21:38
>> They they adapted it into a movie.
1:21:38
The movie's pretty good, but the book's probably better. Have you read the book?
1:21:42
>> I did read the book at one point.
1:21:44
>> Yeah, the book's great.
1:21:44
Um, but 42, of course, is the answer to the universe, all things, everything.
1:21:48
It's a it's a fun fun reference.
1:21:50
Um, the Trump administration will offer artificial intelligence models from Elon Musk's XAI to federal agencies through a new partnership under the agreement uh with the GSA which oversees technology procurement.
1:22:02
Agencies will get access to models such as GRO 4 and a new fast version called GRO 4 fast.
1:22:06
Uh, the deal follows a similar agreement with the other folks in the space.
1:22:10
Automating state government processes, one of the most promising applications for technology, fueling the battle to be the most popular popular tool for federal workers.
1:22:20
All four of the AI companies also each have $200 million contracts with the defense department.
1:22:25
Sort of interesting that uh >> four Yeah.
1:22:30
So, the government clearly wants an igopoly here and is like, we're not going to really pick winners.
1:22:35
which is crazy because Daario has been very outspoken about the Trump administration and Elon's obviously been fully in the Trump administration's camp.
1:22:44
Google has been a little bit quieter and has leaned left in the past. Open AAI.
1:22:49
Sam Alman was kind of stayed out of politics but then has been at dinners and was at the was at the inauguration.
1:22:57
And so, um, you would expect based on like the news of how close the Foundation Model Lab CEOs are to Trump, like just how many photo ops they've been, how many how close have they been, uh, you would expect a massive contract for Elon and a very small contract for Anthropic, and yet they all got $200 million equally.
1:23:14
It's interesting, just like they're they're clearly focused on on having balance amongst the AI foundation.
1:23:21
I want to see the usage >> in the in the in the government broadly and in the DoD.
1:23:28
>> What are people using?
1:23:28
I mean, >> everybody gets a $200 million contract.
1:23:31
It's unlikely that that they're going to be like super evenly leveraged, right? >> Yeah.
1:23:37
I wonder if they would use them for specific things.
1:23:38
I mean, you would hope that like, you know, they're using Claude Code over there for something and then they're using Chachi PT for knowledge retrieval and stuff and then they're using Gemini for the things that Gemini is great. Nano banana.
1:23:47
I guess the government is just, hey, you don't need to come in for DMV photos anymore.
1:23:53
>> $200 million to take great beams.
1:23:55
>> For gray beams, >> great memes. >> Oh, great memes. Yes, exactly.
1:23:57
Uh, but uh you would hope that uh you'd hope that the government's using AI all over the place and they need an actual contract to make sure that it's provisioned and hosted in the proper way. Anyway, Bin.
1:24:08
AI, the number one AI agent for customer service, number one in performance benchmarks, number one in competitive bakeoffs, number one ranking on G2.
1:24:16
You can start a free trial.
1:24:16
Um, Lulu is chiming in about Rick Rubin.
1:24:18
She says, "I mean, just look at them. Can't media train this."
1:24:24
Austin says, "The Beasty Boys, Ed Sheeran, Lady Gaga, Red Hot Chili Peppers, The Strokes.
1:24:29
Rick Rubin has worked with all of them."
1:24:32
>> Why is no one talking about Rick Rubin?
1:24:34
>> Why is no one talking about Rick Rubin?
1:24:36
Um, he's worked with all of them and is considered one of the greatest producers of all time.
1:24:39
His process to creativity can be described as looking for clues.
1:24:44
He has such a unique aesthetic with that massive >> always has the red the the blue blockers on.
1:24:50
>> Those are blue blockers. >> Yeah.
1:24:52
>> Oh, I feel like blue blockers are normally orange orange.
1:24:54
But I guess those are just a dark orange that I'm seeing the lighting shade. Yeah. >> Okay.
1:24:59
>> Um well >> um >> what was else is going on?
1:25:06
>> Scott, >> your post here.
1:25:07
Uh there were some accusations accusations earlier that somebody was doing a paid post. >> The post was deleted.
1:25:13
So, it was uh Mike Isaac at the New York Times picked up a device called a brick, which is a little NFC chip that they see, you've se everyone's seen this on Instagram, right, where uh it's supposed to like brick your phone so that you don't be distracted.
1:25:30
And so, you have the willpower just once to attach this to the back of your phone.
1:25:36
It kind of magnetically attaches and it puts your phone in a do not disturb, don't show me any of the addicting social media apps, turn off everything, let me focus.
1:25:45
Um, >> let me cook, >> let me cook, let me lock in.
1:25:49
And so, uh, apparently, according to the original post, Mike Isaac picked one up off of an Instagram ad and said he actually loves it. It's helping him write.
1:25:59
Of course, he writes for the New York Times.
1:26:00
He needs to lock in and write a lot. He can't be distracted.
1:26:03
He's also a prolific poster.
1:26:05
can't be caught on the timeline mid article when he's on deadline. >> That's right.
1:26:10
>> So, uh he posted about it and uh a uh >> another poster >> another poster accused him of doing a sponsored post without disclosing it. >> Undisclosed ad.
1:26:20
>> And if you know the New York Times standards and who Mike Isaac is, he's published a book that got turned into a movie or TV show.
1:26:27
like he's probably not doing Spawn Con for some DTOC like you know electronics company undisclosed.
1:26:37
It's just like so much risk to so little reward like like the post that of him saying like oh I got this brick and it's great like it didn't get that many views.
1:26:46
It's not like brick would be like oh yeah we're going to sell 10,000 of these things off of this.
1:26:50
Uh, let's pay Mike Isaac $100,000 to post this, you know, organic content, this fake content.
1:26:58
Uh, >> he just he likes technology.
1:27:01
>> Yeah, he likes technology.
1:27:01
People like to post things that they test out. >> Covers consumer tech.
1:27:06
>> Not not everything is spawn con, >> but I thought your post was hilarious.
1:27:09
You said, "It's time to come clean."
1:27:10
Responding to this allegation.
1:27:10
Uh, John said, "You need to admit that Brick paid you to dunk on this organic post to drive more impressions.
1:27:17
This post is brought to you by Brick."
1:27:19
I do think there's something funny where you could potentially run some sort of like anti-astroturfing campaign as a brand where you pay a bunch of of influencers to accuse organic UGC of being paid and it drums up way more attention.
1:27:37
Uh and and so I was riffing on that.
1:27:40
Uh but on this show we don't do undisclosed partnerships.
1:27:44
We do disclose partnerships like our partnership with Adio customer relationship magic.
1:27:50
Adio is the AI native CRM that builds, scales, and grows your company to the next level. Get started for free.
1:27:55
When we read an ad, you know, we're getting paid for that.
1:27:58
You know, we're not doing it for free.
1:28:00
We're not doing it for the love of the game.
1:28:01
We're doing it for >> Well, we are doing it for the love of the game.
1:28:03
Nobody loves SAS more than us. >> That's true. That's true. Um, let's keep going.
1:28:10
There is a lot of stuff going on.
1:28:13
Um, this was a really funny post because it gave me a total jump scare.
1:28:16
I was talking to Brandon about it earlier.
1:28:18
Uh, Ben, uh, Podgi, sounds like a podcaster's name potentially.
1:28:23
Uh, Ben says, "This is correct, but we should take it to the natural conclusion."
1:28:28
He was, I guess he was quote tweeting Rune, which we'll get into.
1:28:32
Um but uh he says uh like the Habsburgs of old the USA should look for synergistic M&A not through immigration but with entire policies.
1:28:43
In this case the obvious answer is merge South Korea in as the 51st to 57th states.
1:28:49
I was like obvious totally obvious >> not the obvious answer to me.
1:28:53
When people talk about, you know, growing the United States, they talk about Greenland, they talk about Canada, Mexico, Puerto Rico, the Fiji, like the US Virgin Islands, like there's so many different territories that are like more closely aligned with the United States than just going all the way over to Asia and just picking up South Korea.
1:29:11
But Ben makes the argument.
1:29:13
He says merging South Korea in would resolve a wide range of defense, industrial, trade, and demographic crises.
1:29:19
South Korea would not have to hem and haw about building their own nuclear shield.
1:29:23
They'd be an inviolent part of the United States from the US from day one would have would have massive ship building capacity vastly improving national defense and US companies could frictionfree contribute in the one spot where we genuinely have a technical contribution small nuclear power systems.
1:29:41
The US's relatively healthy demographics could gradually backfill South Korea's catastrophic demographics.
1:29:47
Uh maybe we could even teach them how to have kids again.
1:29:49
The free flow of labor would allow Korean workers to train US workers without insane visa shenanigans.
1:29:54
South Korea would neither uh is neither unduly liberal or conser or conservative relative to the United States.
1:30:02
um small couple uh I like the thought exercise.
1:30:08
Couple small potential problems here would place the United States directly on having a shared border with North Korea and be right across the Yellow Sea from China.
1:30:17
Uh >> it is such a wild move, but he lays out a, you know, a thoughtful case. >> It's a cool idea.
1:30:25
I mean, maybe it's no coincidence that uh the most respectful TBPN uh derivative show inspired show is from Korea. >> That's right. Those guys know do it.
1:30:37
>> Also, I love uh Have you ever had BIM beam bop?
1:30:40
>> No, I don't know what that is.
1:30:41
>> Popular dish in South Korea.
1:30:41
It's like rice and protein.
1:30:43
It's >> Have you ever been to South Korea? >> I have not.
1:30:47
>> I've never been there. Have you?
1:30:48
>> Uh have a >> Have you Have you ever traveled anywhere in the world? >> Yeah. >> Where? >> I've been to Ireland. Oh, okay. Job's finished. Moving on. Ireland world traveler. >> Luck of the Irish.
1:31:04
>> So, this is all all kicked off by a rune post.
1:31:07
He said, correct me if I'm wrong, but it seems like the theme of the Dan Wang book who we had on the show.
1:31:09
And the general elite consensus now is that industrial process is a technology that lives in the heads of people and that it was a mistake to let so much quote unquote lowv value industry be offshored due to the tacet loss of process capital.
1:31:23
TSMC Arizona which makes the most complex and valuable industrial production in the world was a massive success.
1:31:30
This was a huge surprise to me.
1:31:32
I did not expect the TSMC buildout.
1:31:35
Everyone was saying like it's impossible to airlift TSMC.
1:31:37
Uh but they wound up producing four nanometer chips at great yields on par with Taiwan merely years after striking ground for the first time.
1:31:47
This involved a generous federal subsidy and importing thousands of the great Taiwanese semi semiconductor experts despite unions trying to quell Ty Taiwanese immigration and some culture clashes in the United States.
1:31:59
Aqua hires of whole teams with process knowledge in their heads is very common.
1:32:03
Zuck acquiring some of the greatest talent from other AI labs for massive numbers is just one example of this.
1:32:09
Also seen in the full self-driving wars between Uber and Google, which was interesting because that is of course about uh the Anthony Lewendowski case that ended in a lawsuit that landed Anthony Lewandowski, I believe, in jail for a little bit, but then he was pardoned and so he got out.
1:32:24
But um this it was more I feel like it was more than just process knowledge in that case because it was specific patents.
1:32:32
And I'm wondering how much how much of a line you can draw between um knowing a specific algorithm knowing a specific something patentable and actual process power process uh uh process knowledge.
1:32:48
But uh it's certainly an interesting analogy to to dig into.
1:32:50
So Tesla plus Apple plus big pharma acquires industrial process companies all the time.
1:32:57
America is very capital rich able to levy literally hundreds of billions of dollars for machine intelligence capex.
1:33:03
We can afford to acquire whole groups of foreign talent for prices that are unheard of to them in their home countries.
1:33:10
Uh TLDDR aqua hiring foreign process knowledge for massive sums should be one of the primary goals of any re-industrialization effort.
1:33:16
Special visa categories should be made for to scoop up whole teams of Senzhen's best.
1:33:24
The raids on the LG battery plants in Georgia are the exact opposite of what we need.
1:33:28
Ability to tolerate new arrivals is a technical edge of American capital to be able to assimilate foreign knowledge into domestic industrial processes at a scale nobody else can countenance.
1:33:38
countenance. I was I was hearing a story about how the I think what was it the Manis team relocated to Singapore and there's a world where they relocate to San Francisco >> set up in the benchmark office >> maybe I mean >> little satellite office >> yeah I mean there's there's there's
1:33:55
something there I don't know how how important their process knowledge is but um certainly it worked for TSMC I don't know how many more of those projects need to be done Taiwan seems to be in a uniquely precarious position Whereas a lot of the South Korea tensions seem to be a lot lower. Like
1:34:11
Like there's less geopolitical risk in South in South Korea.
1:34:15
So there's less of a we need to move SK high to America.
1:34:18
Uh but Tyler, do you have any thoughts on this post? I'm sure you saw it. Um what do you think?
1:34:25
>> Yeah, I think it makes a lot of sense.
1:34:25
I think uh politically it might be hard to do this, but I've seen a lot of similar arguments um just for like normal AI researchers in China.
1:34:33
We should basically have some kind of like visa that just like imports them like instantly and then it's like the I I assume the salaries over there are not comparable to like MSL level like $100 million. >> Totally.
1:34:45
This is the real This is the real AI paper clipping.
1:34:48
It's not that you get turned into a paperclipip.
1:34:49
It's that you get operation paperclipipped into a different country and we need to avoid our best getting paperclipipped to a different country and we need to be potentially paper clipping other uh national champions.
1:35:02
I wonder if there's something to do in uh in in AI in not just AI but in uh the process power around solar because we've talked to Casey Hammer about this.
1:35:12
It feels like the vast majority of the cheapest solar panels come from China.
1:35:15
We're maybe not we don't have a national champion there yet.
1:35:19
We're not catching up but uh fortunately we have a guest who might be able to get us up to speed on this.
1:35:23
We have Delhi Naspuhof from Founders Fund.
1:35:27
He is back from paternity leave.
1:35:27
He's in the reream waiting room and he will be joining us in the TVPN Ultra Dome in just a minute. Let's bring in Delian. How you doing? >> Yes. >> Good to see you. >> What a voice. >> Daddy's home. >> Daddy's home, baby. >> Oh, there we go. >> Got my swag. >> I love the hat.
1:35:46
How How is it being a uh a father of multiple now?
1:35:52
Um, you know, I feel like you blow up your life once and then after that you kind of you guys sort of figure it out where it's just like, yeah, you got the infrastructure and that's why God gave you multiple hands is, you know, one baby per hand. >> Yes. But everything's good.
1:36:02
Everyone's happy and healthy.
1:36:04
Everyone's back to normal. >> Oh, yeah. Oh, yeah.
1:36:08
I mean, she's still only like, you know, sort of month old, so, you know, still requires, uh, you know, some work, but yeah, in the grand scheme of things, happy, healthy, very grateful that, uh, you know, all things pretty smooth. >> Yeah.
1:36:17
Well, once once uh once she gets to two months, it's, you know, it's it's cake. put her to work.
1:36:21
She can come on She can come on TBNN.
1:36:24
>> We've got We've We've had three members of your household live on the show. We need the fourth. That's true.
1:36:29
Well, actually, I mean, your brother, too. So, >> that's true. Five.
1:36:33
>> The Aspar Ruhov clan is dominating.
1:36:36
>> I want at least 10 Asparohs on TV by the end of the decade. >> Yeah.
1:36:39
What What What are your parents up to next week? We'll get them on.
1:36:41
Um, uh, really quickly, I I want to get a general update on what's going on in your world, but, uh, I would be interested to hear your response to this, uh, this debate that was going on on the on the timeline yesterday from Rune about, uh, TSMC Arizona being a good, uh, a good example of airlifting industrial process from one place to another.
1:37:06
uh he kind of ties it to the acquisition of whole teams at MSL uh America having being capital rich being able to actually afford to go and say here's a hundred million dollar offer to some great researcher come to America potentially and it gets into the obviously there's a whole bunch of immigration stuff going on and cultural stuff but did you have a reaction to that?
1:37:28
Are there other industries where it seems like we need sort of a TSMC Arizona like project where we're just going to an international company and saying do what you do over there here. >> Yeah.
1:37:41
Um I feel like a lot of the reason that people are discussing this so much on the timeline lately is because of Dan Wang's uh new book Breakneck where you know breaks down basically the like at least his um argument is that you know uh China is the engineering society where the you sort of lawyer society and everything basically industry stems from there.
1:37:57
Um which I think is probably a little bit you know I like a lot of parts of his book that actually particular argument is probably one of my like less favorite you know sort of parts.
1:38:07
partic um there are things that China is obviously you know sort of very good at.
1:38:12
I think I was you know hearing your last guest talking about like you know mass production of solar panels.
1:38:16
They're obviously like phenomenal at that type of like commodity cost curve relatively complex but not deeply complex type of manufacturing right like they've stepped up from like you know the early days of Shanghai being you know sort of toys and knick-knacks etc.
1:38:31
Obviously now into like you know mass consumer electronics unit is obviously like by far the best you know sort of humanoid robotics company.
1:38:35
humanoid robotics company. Um and then obviously in solar panels um but there's you know I think some limitations to their approach like you know when I think about it from like the aerospace perspective look a Falcon 9 rocket landed for the first time a decade ago now the Chinese basically like stole the IP and plants that and clearly have something that is like the equivalent to a Falcon 9 and they're still basically
1:38:55
doing the like early days of like you know Star Hopper which was you know back or Grasshopper I mean which is like the SpaceX project in like 2013 I want to say um that was just doing the suborbital little like hop ups and downs to practice for Falcon And so I think I do think they like have this limitation when it comes to like deep deep systems engineering and requiring some creativity that they like miss out on. And I think that's why
1:39:17
And I think that's why you've seen them partially succeed in semis right they're good at the like you know sort of last generation of semis and you know sort of reshoring that to China but I don't think you've really seen any progress on them you know sort of you know domesticating cutting edge you know sort of semis.
1:39:31
Um, TSMC Arizona, you know, is at least a very subscale version, you know, of um, you know, being of being able to do actually some amount of cutting edge domestically.
1:39:40
Um, and I do think it, you know, sort of shows that there's just some breakdowns in Dan Wing's argument where like we are able to import that type of like I forget what he calls it like industrial process or process knowledge um, in the United States and even develop it ourselves, right?
1:39:52
The other area that obviously like we do quite well on, you know, it's not the perfect company, but like look like the, you know, 99% of commercial ainers in the world are still developed by the West, right?
1:40:00
you know, Boeing and Airbus and continue to be um and it's not obvious that China has really been able to displace that.
1:40:05
Um you know, the only other areas that I think about actively that you need to be, you know, sort of reshored some amount of it is like, hey, if we want to, you know, to compete against China and Taiwan, it is going to be almost certainly like a mass scale production, you know, sort of game more so than anything else.
1:40:21
And so when you think about like mass scale drone manufacturing, mass scale metal cutting, mass scale like fighter jet production, that's something that like we're obviously behind on.
1:40:29
even like the munitions, right?
1:40:30
Like, you know, Ukraine has shown that like, you know, artillery is back to being relevant again.
1:40:35
It was not relevant at all in any of the like Middle East conflicts, but now with like but now with like continuous, you know, internet coverage and continuous basically, you know, spy satellites, you know, you can actually justify basically just like, you know, anytime that you're, you know, enemy makes a move, you just artillery shell the crap out of them.
1:40:50
And that's actually like back to being, you know, sort of relevant on the battlefield.
1:40:52
Um, and they are far better.
1:40:54
or they have these like, you know, lights off, you know, um, artillery shell production facilities that basically just like are 24/7 fully roboticized, you know, produce, you know, way more than the United States can.
1:41:04
And so, um, I don't know that those are going to be necessarily like the equivalent of TSMC where you're like taking somebody's pre-existing industry and like dropping it here and that like there's not an obvious place to go import from basically for that type of stuff.
1:41:14
So, I think we kind of solve that on our own and obviously founders fund were making some investments you know in that space obviously. Yeah.
1:41:21
>> Isn't it true too that in China like yes there there's flagship lights out factories that are truly cutting edge the kind of thing we just don't really have here in the United States but yet still the vast majority of manufacturing in China is just a lot of labor like you
1:41:39
know assembling cheap components >> it's shifted I think that's like a you know sort of decade out or a decade old sort of you know sort of thinking where it's like yeah the reason that China won for a long time was because of just like the labor arm, a cheap cost of labor when you talk to people. >> But but I'm just saying like with with
1:41:54
>> But but I'm just saying like with with the true like we can't buy the like when you talk about China's advantage of just being able to produce overwhelming mass right across every different sector.
1:42:05
across every different sector. There's certain key areas where they are just light years ahead and then it's they're still getting this massive benefit of even the components that get made that go to the lights out factory to be assembled you know but with robotics like there's >> yeah I mean when I think about like
1:42:24
their advantages in like you know drone manufacturing is mostly due to automated facilities right like it's not because they like you know have a bunch of like you know sort of mass assembly super precise small hand you know sort of laborers or something like that It actually is because when you think about like the PCB boards largely basically fully
1:42:41
automated the like you know propellers that they make largely fully automated the battery lines largely fully automated and so yeah I I just like I when I think about like the edge that they have in drones yeah there's maybe some amount that's like you know sort of this like assembly edge but like I put that at like 5 to 10% of the edge like
1:42:56
90% of the edge is that they like have such deep process knowledge across all these individual like different you know basically like subcomponent sectors and a lot of those are fully say again co like collocation effectively put >> yeah and the collocation is just you know sort of wild right and this is also the sometimes the problem with like the
1:43:11
stuff in America when we like try and reimp import it it's like you know the chips act one of the biggest battles was like the you know the size of funding required support from a broad set of senators and congressmen but then by default they want you know sort of jobs everywhere and so it's been more difficult to fully you know sort of
1:43:26
centralize all of the like you know reshoring of semis in the United States into you know sort of single location like you know Arizona's definitely you know sort of been impressive but like ideally it's not just the like fab at some point you want to do like lithography there and a bunch of other parts of semis centralized there but like in the United States we actually
1:43:41
kind of have a bit of a chips hub obviously still some amount in Silicon Valley Austin's become like a you know sort of really big area ideally those would all be in the sort of the same place rather than so spread out and so it is just harder given that we're not like China that's willing to sort of like put their finger on a scale to a
1:43:55
single you know city and just be like you are going to become like the guitar producing city and everything is going to be you know sort of based there we don't we don't do that >> need a new guitar executive It does feel like we're starting to do that with the Neoclouds, with some of the big data center buildouts, Abalene, Texas. Like
1:44:09
Like we're we're marshalling capital, >> but that's still led by individual dealmakers going and talking to these local governments and figuring out >> Yeah. >> Can we do this here? Okay.
1:44:21
No, I'm going to go across state line over here. I'll do it over here. >> Sure. They're chasing. >> Yeah.
1:44:27
We're just much less top down coordinated, but like at the same time, I think you're starting to see like the faults and the breaks in that system, you know, in China.
1:44:33
And Dan Wing talks about this as well where it's like look, you know, they clearly still excel in certain areas, but they've also like I mean like think about it even over the course of time that I've been at Founders Fund in 20121 like the number of unicorns being minted in China was starting to approach, you know, the United States and there's like some real risk and fear that it was going to supersede.
1:44:50
Now four and a half years later, you know, the Chinese venture capital ecosystem is effectively, you know, totally negligible.
1:44:55
And so um you know I do think that they've you know done a decent job on like you know sort of prioritizing the industries that matter the most but also their like capital markets are like you know sort of way worse than they were you know sort of four and a half years ago.
1:45:07
Um you know they have reimported a bunch of you know sort of Chinese nationals from the United States back as scientists into China you know pretty good but they've also like lost you know something on the order of I forget what the number was like 25,000 Chinese millionaires like are you know exporting their capital.
1:45:20
>> They lost the Manis team.
1:45:20
The Manis team's over in Singapore now.
1:45:22
They might be working out of the benchmark office soon. Yeah.
1:45:26
>> Well, speak speaking of benchmark.
1:45:26
I thought it was interesting actually Bill Gurley posted earlier game recognized game and it was a quote from the CEO of Xiaomi who said, "We bought three Tesla Model Y's for disassembly and research inside Xiaomi earlier this year. What a great vehicle."
1:45:41
And of course, they're announcing like effectively a copycat, but it's insane to just say >> game recognized game. It's game copies game. >> Yeah.
1:45:50
Just it's saying saying the quiet part out loud is wild.
1:45:52
part out loud is wild. But uh >> recognize I think by the way since I was last I was last on was when I have been poking the benchmark bear literally basically like since like you know a week or two after joining Founders Fund and I remember early on I got a call from like one you know one of the partners at FF being like hey like look
1:46:09
just so you know you're definitely making some enemies over there and I was like okay and then they were like but I wouldn't necessarily like it doesn't feel right to tell you to like take these things down because like at the end of the day we are founders fund and you are critiquing them for firing founders and that is our whole ethos. So
1:46:21
So just know that what you're doing is definitely creating enemies and you are creating risk.
1:46:25
So anyways, I've been poking the bear for like 5 and a half years.
1:46:28
And then it was so glorious to have like finally an official response from Cheton.
1:46:34
And it was just like it was an alleyoop.
1:46:36
It was the perfect response.
1:46:38
It was like how dare you not basically like you know you know control you know sort of a company's choice in you know sort of you know capital provider like why are you letting them take Chinese capital?
1:46:46
And I was like oh my god this is perfect.
1:46:47
my god this is perfect. is like they exactly you don't even realize how wrong you are because you don't even realize that what you're saying is why aren't you an adventure investor telling the founder how to run their company and that's your critique and I remember reading it and I was technically supposed to be on this like oneweek like
1:47:01
Twitter whatever like you know detox um but then like you know >> you're on on Twitter being like I'm supposed to be on a detox right now but I'm about to go on a bender >> I know but I was like I got like five texts from friends being like how have you not responded to this yet and I read it and I was like oh my god yeah this detox bro definitely going in. And it was just like it is I think
1:47:19
And it was just like it is I think either my top or second best day on Twitter in terms of joy provided for me like like true like existential like identity level like as much as like a child's birth.
1:47:32
I mean like that is the level of joy >> zero user regretted seconds like the least the least regret of >> you would compare it to Yeah. holding your daughter.
1:47:44
>> It is like it is Yeah.
1:47:44
I think roughly equivalent.
1:47:46
The only other day that I've had that was that good was the like the day after the FTX blow up and I was the first one to really find and screenshot the Sequoia blog post about him playing League of Legends and getting that out and it was just that was just that was incredible.
1:48:03
It was just it was it was like I got to break news.
1:48:05
I got to like poke fun at a competitor.
1:48:07
I forced them to like have a response to all their LPs about it.
1:48:11
They had to like take down that blog post.
1:48:12
It was just like that was like a truly >> um weren't you also better than SPF at at uh League of Legends? >> Okay.
1:48:22
I mean like AOC is better than Yeah.
1:48:26
at like League of Legends.
1:48:26
So like the bar is very low.
1:48:27
I just like >> But at least at least you got to respect some of uh some the the highly convicted investments in anthropic.
1:48:34
Robin Hood, you know, he was a gambler, but you know. >> Yeah, >> dude. He was a good gambler.
1:48:40
I mean like he just you know played the line a little too hard but like yeah there's a world where like yeah I forget who was telling me this it was somebody that was basically responsible for like brokering the like FTX like bankruptcy etc.
1:48:48
So they're like a you know Wall Street whatever you know sort of type and they basically were like yeah like if he had just lasted another like 6 months the book like would have been totally fine.
1:48:57
He wouldn't have got liquidated and today if he just held these investments that he made from like back then he would be richer than Elon.
1:49:02
Like he would actually just be the richest person on the planet. >> Wait what?
1:49:05
>> And so it's just it's wild to think that he was like on a knife's edge >> richer than Elon.
1:49:09
I don't how's he putting up >> I I haven't seen that actual number but I mean I guess it makes sense with like the various positions he had.
1:49:15
Yeah, >> I think it was like assuming that FTX continued to go eternal >> there were like I mean again like some of those you know moves were just like like it's like the anthropic position I forget is like now like $25 billion.
1:49:28
>> I mean the guy was insane.
1:49:28
Yeah, I I don't know if you remember, but he went on Tyler Cowan's show, Conversations with Tyler, and Tyler asks him, "Would you flip a coin where there was a 50% chance of humanity being completely wiped out or 50% chance that you double the prosperity of humanity, but it's 5149, the good outcome?"
1:49:47
And he was like, I would flip it a hundred times in a row.
1:49:51
And it's like, that's insane.
1:49:54
Like, like that's not how these things work.
1:49:56
But he was like, "The expected value is positive, so I must take the the EV positive bet."
1:49:58
And it's like, "No, no, no, no, no.
1:50:00
Please do not flip that coin.
1:50:02
Like, you're going to wipe us all out with like 99%."
1:50:06
>> The best part of SPF having to go to jail is Dustin Mosov is finally shutting the up about EA. Thank God.
1:50:10
We can stop talking about only net EV positive things. >> Yeah, it's wild.
1:50:16
Well, uh, speaking of net EV positive things, is it no longer net EV positive to invest in a company that's growing revenue triple, triple, double, double, double?
1:50:26
Because I have a lot of friends that have been doing that.
1:50:29
They previously they were say, I'm about to IPO my company, but now they're thinking of shutting down because >> according to the best venture capitalist, >> Delian's a fan of 0000,000x.
1:50:41
>> But yes, I want your reaction to the state of the software markets.
1:50:43
It's sort of an EV take a little bit.
1:50:47
>> Have you have you done a pure SAS deal this year?
1:50:51
>> Um uh I did just do like a like basically pure vertical AI SAS stage even had to do one he had to do >> vertical SAS underrated criminally underrated.
1:51:06
>> That's you literally talk to every VC.
1:51:08
It's like you you just go go invest in the foundation model layer. Go invest in hard tech. Leave this.
1:51:13
Leave the vertical SAS for for me. >> That's great. >> Yeah.
1:51:19
>> But yeah, overall thoughts on the triple triple double double double being being like out of out of like out of date with how fast the foundation model companies are growing.
1:51:27
I had this interesting thought that it's like you could be running a janitorial company and if you have a foundation model as a client and they're growing 10x, you're going to be cleaning 10 times as many toilets.
1:51:37
Your revenue is going to be 10xing because they're 10xing.
1:51:40
And if you just like are a barnacle on the side of the massive whale, you're going to see massive growth.
1:51:45
And I wonder if like that like what what that means because >> we're investing in the barnacle economy. >> The barnacle economy.
1:51:51
Uh but in general like you know what what are your thoughts?
1:51:55
thoughts? Yeah, I think about this stuff all the time from like the you know sort of space perspective too where like you know you can get delude into this world where it's like the space economy is growing but if it's only a bunch of other space startups like you know trading revenue back and forth you can like prop that up for like quite some time trading a dollar back and forth but at some point it needs to connect into
1:52:10
like the rest of the ecosystem right and so >> you know I think there was you know some you know memes and jokes going around you know yesterday about the like Nvidia Oracle OpenAI whatever 100 billion investment etc and it's literally just like the same dollar effectively going you know sort of in circles over and over again Um, and that can, you know, sort of work, but like that can also blow up. And we just, and I think we
1:52:28
And we just, and I think we talked about this in maybe even in the last interview.
1:52:30
It's like you have just this crazy dynamic where, you know, the percent of GDP being invested in this category is like, you know, the equivalent of like at the railroad bubble, etc.
1:52:39
>> But a lot of the companies that are doing this have like these cash flows from other very large businesses that are driving towards this, right?
1:52:44
You know, Zuckerberg, I think, gave the quote where he's like, "Look, I would prefer to like, you know, waste hundreds of billions of dollars, >> misspend a couple hundred billion." >> Yeah.
1:52:52
spend a couple hundred billion rather than lose the machine god race.
1:52:55
And so it's just like the level of it just it's like nothing we've ever seen before in terms of like the super cycle of you know sort of capital going into this.
1:53:02
Maybe it has the ROI maybe it doesn't.
1:53:05
When I think about it from like the investor perspective I mostly think about it from twofold which is like how you should analyze any business which is basically like quality of revenue and like you know you know in terms of like durability of that revenue and then you know the competition eg the margins you know sort of on that revenue.
1:53:17
And so I'll maybe provide like one example that I think is very strong and one that I think is like a little weaker.
1:53:22
Um you guys have had you know sort of Melissa um you know John Lud you sort of wife on for cybernetic labs.
1:53:27
I think of that as a great example where you know she has really great revenue growth.
1:53:31
really great revenue growth. Um and uh uh I believe it to be very durable in that most of her revenue growth is coming from things that are totally decoupled from the like insanity of like you know Nvidia OpenAI etc circles like it literally is janitorial companies
1:53:47
right effectively HVAC plumbers etc >> those people's revenue their growth etc like is completely like economically decoupled and so in some ways then Melissa's revenue is completely economically decoupled from the entire AI hype wave and so when I think about like oh okay let's say imagine a world
1:54:02
where all of a sudden like you know the mag 7 decide to like you know reduce deployment into data centers and like foundation model training by 100x you know next year which companies are affected by that okay well like coreweave I imagine is like going to be affected by that Oracle definitely going
1:54:16
to be affected by that openai and thrope definitely going to be you know affected by that >> yeah you look at you look at how many companies are now indexed to open aai right you have broadcom oracle soft bankank coreweave like these are >> there's a lot of them Yeah, a lot of them. >> And I think as an investor, you have to
1:54:33
>> And I think as an investor, you have to take a little bit of like a bifurcated strategy where it's like it's I think it's a generational index.
1:54:37
And this is, you know, if there's one thing that I admire about Peter the most is, um, he just does such a great job of even though he has lots of biases, particular, you know, um, preferences, etc.
1:54:50
on one sector or another found or another.
1:54:52
another. He's very willing to very I feel like soberly analyze the just like broad macro world and identify hey even if like I don't love you know whatever boring SAS software etc or like the AI stuff you know you may not be real still be willing to like analyze it from a
1:55:06
purely investment perspective and like take it on and obviously we've you know done decent you sort of checks into this field versus like I literally like find my biases so strong that I can't even like I can't even motivate myself to even spend time on it you know to like you know analyze it as an investor. Um
1:55:17
you know analyze it as an investor. Um so that's where I think about the you know sort of you know durability perspective and then there's like the margin you know sort of perspective which is like how much you know competition is there right you're seeing obviously in these foundation models where just like the token cost continues
1:55:29
to go down the margin never seems to really be improving and a part of like there's a bunch of you know companies that and capital that's going into this and you know some of the like let's say like AI for you know whatever you know coding has maybe a bit of a similar dynamic where it is like you know a bunch of companies that are working on it the foundation models working on it etc. I don't think there's going to be
1:55:47
I don't think there's going to be like 15 AI for plumber, you know, sort of companies, right?
1:55:50
And so this is where Melissa both has this like great durability and great margins.
1:55:53
And so I think you have to take this kind of bifurcated approach as an investor which is both like invest into the like index effectively.
1:56:00
effectively. Now our approach has generally been invest mostly into the thing driving the index eg openi um and then you're like you know sort of counter strategy is invest in things that are totally decoupled from the index and like super far away and then basically like nothing you know that much in like the you know sort of messy
1:56:15
middle in between and to go back to like the you know whatever you 332 or whatever it was um I do think that in both categories yeah I do think there's some merit to it where it's like the growth rates and EV growth in these companies just do look very fundamentally different and the bar is definitely higher. It's just been like I
1:56:33
definitely higher. It's just been like I like it's been very remarkable to watch Melissa and her company grow at a rate that just like yeah I I just like I you know it if you had told me you know write the distribution of where you think Melissa's revenue growth is going to be like when she like started the
1:56:46
company this is in like the 99th percentile you know and and I think it just speaks to how much of a wave this can be and I think Peter talks about this where it's like the internet wave happened but then it still took like 20 years to get the integr internet integrated into the rest of the economy. I think you're going to see the same
1:56:59
I think you're going to see the same thing, you know, sort of here where like there might be not as bad of like an AI bubble crash, but the like implementation period still may take might be more accelerated because just generally society is moving faster, but it may take 10 years.
1:57:09
And it's companies like Melissa that are getting it integrated into like the broader broader society.
1:57:14
>> Have you seen any companies where they've done a good job?
1:57:17
They're still in founder mode.
1:57:18
their uh their they they've grown like a base of business that then when the AI story came they could actually act they could act on that in a way that it wasn't like you know they they're a public company and they need to go through this massive transformation of how they build but they can just add AI on top and then they're they're actually growing new quality revenue.
1:57:41
Are you uh are you feeling that at all?
1:57:44
What are you pointing to?
1:57:46
>> Let me figure this out.
1:57:46
Sorry, I'm having to >> Oh, yes, yes, yes. Ramp. >> There we go. There you go. >> Yes. Yes. >> There we go. >> There you go. >> Ramp. com.
1:57:56
Um, you know, great company that has done a, you know, great job of like, look, they obviously like were growing very well before any of the, you know, the AI boom.
1:58:05
>> But like I just remember this very particular, I think they've announced this now, so I'm not like revealing anything, right?
1:58:08
anything, right? they have their like you know AI agent you know basically product but I just remember the board meeting where Kareem and crew presented like the vision for it >> and I was like oh this is like what a very competent but like larger startup that is still fast moving knows how to
1:58:23
adopt things does in this world whereas basically like yeah like look we have a bunch of people that obviously use what is a very like um beautiful interface that automates a lot of like you know finance work but it still requires people to go in set up these rules engines do some amount of tagging etc. we're just going to like basically start
1:58:38
we're just going to like basically start to offer for you know some of our beta customers that are going to try this out.
1:58:43
We're just going to like effectively screen record what they're doing and start training AIs on that.
1:58:47
And it's like I just can't imagine like both if you're a net new company being like I'm going to make AI agents for the CFO suite.
1:58:53
It's like well but there's like a whole set of things you have to do.
1:58:55
You have to have like the corporate card.
1:58:56
You have to have travel expenses.
1:58:57
You have to this expense policy accounting integrations into the ERPs integrations into like their tax.
1:59:01
It's like >> all these systems that you need to build that are like more web 2.
1:59:04
that are like more web 2.0 O whatever you know sort of systems and only when you have that and you have a ton of usage on it >> like can you then actually start to train the like AI agents on top of that usage right and so I do really I've admired that a lot in ramp and so that's probably the one portfolio company that I've seen the strongest in and then the
1:59:21
second is this other portfolio company of mine sword health that um you had been doing this like uh basically you know >> AI sort of physical you know sort of therapist but most of it was that basically you had these like sensors that were strapped on your body you would go through your physical therapy routine and then a like PT would basically then review your movement data set, etc., and like message back and
1:59:37
, and like message back and forth with you and adjust it.
1:59:39
That was a very obvious area where like they literally had years and years of training data sets on how their digital PTs were basically interfacing with patients.
1:59:47
They basically like significantly up the ratio.
1:59:48
I forget the exact number, but like my guess would be it was something on the order of like I remember when the company started it was like 12 patients per PT.
1:59:52
I remember by like 2022 was like 1 to 100 and I think now it's something on the order of like 1 to a,000.
1:59:58
And it is largely like the like you know AI you know sort of wave that has enabled them to do that and especially because it's like this text back and forth literally with like a you know sort of physical therapist.
2:00:06
But again if you were starting from scratch be like I'm going to make AI agents for physical therapy.
2:00:10
It's like kind of hard to do like just that >> the physical sensors the distribution with like Fortune 500s health systems etc. >> Yeah.
2:00:18
We had a company on I think a couple days ago called Filevine that raised $400 million.
2:00:22
They were for 10 years creating just workflows for different law firms. >> SAS for law firms. >> SAS for law firms.
2:00:29
They have all the workflows, time tracking, conflict checks, etc. document storage.
2:00:33
Now they can just add AI in a bunch of different places and maybe they don't have their own foundation model, but like maybe claude or open AI models become good enough, right? >> Yeah.
2:00:47
give us an update on uh on space defense tech stuff that's happening in DC.
2:00:53
It seems like there's a ton of deals there, but they're two orders of magnitude smaller than the AI sometime.
2:01:00
It's like a billion dollar deal for Palunteer in the UK and people are like uh I I just heard about a hundred billion dollar deal over here from your video, but uh there's some amazing stuff happening.
2:01:08
What's been top of mind for you in space or defense or DC generally?
2:01:12
Yeah, I mean the you know sort of big macro story for the next five days is it seems like we're very likely headed to towards a government shutdown.
2:01:17
Um I think we've you talked about this in some of prior you know sort of appearances. Yeah.
2:01:21
Uh there's going to be a lot of people for a load rip.
2:01:25
>> Um you know we've talked about this in prior I think >> can you imagine if we had a a precedent like that in the private markets where like companies just routinely like yeah we're out of out of cash.
2:01:35
We're talking with the board but >> everyone's fured. >> Everyone's furoughed.
2:01:39
>> Everyone's furoughed. just probably it'll be like 5 days maybe like 40 days you know whatever >> imagine was imagine like >> Xiinping coming out and be like oh yeah sorry like we couldn't like you know with my finance minister agree on a budget and so we're shutting down the
2:01:52
CCP for like you know it's just like it's so funny to me that like they so badly want to assurp America and our system still has us regularly shutting down for extended periods of time unable to the most basic facts of like what should we be even like prioritizing and allocating towards and We still kick their ass. We have the best capital
2:02:10
We have the best capital markets.
2:02:12
We have the best nation, the best freedom, the best technology in almost every single area.
2:02:16
>> If you're a career politician, it must feel so good to shut the government down. >> Oh my god.
2:02:22
I just what I feel bad is like now that I we're so like so deeply integrated in all these government programs is like there's a lot of people that just like end up having to like continue to do their job, but like they just don't get paychecks.
2:02:31
And it's like and some of these people are like friends of mine now that like have worked in technology and are over there and it's like yeah like you just kind of have to make sure that you have savings because like rent still keeps going.
2:02:40
You're kind of still expected to work but you literally just like don't get paychecks and at some point you get some amount of back pay.
2:02:44
But um but yeah the reason we're marching towards it is like look we've gotten into this world where like >> you know I think it's something like four of the last 20 years we've actually passed like our budget on time.
2:02:53
So the fiscal year for the government basically ends you know basically end of September every year.
2:02:57
every year. um we effectively for many many of the last 20 years we never end up you know basically actually figuring out what the budget should be so we go into these continuing resolutions but when you're in that state and you're
2:03:07
doing everything so last minute it means much more regularly we have shutdowns so shutdowns are happening you know sort of more often and so this year it's the Dems and Republicans going back and forth on like you know basically how much should be based just off the big
2:03:17
beautiful bill you know how much you know should the you know sort of dems have to you know sort of fold on you know some of their asks around you know a lot of it is like healthcare spending you know basically related >> and at least right now the like tone in
2:03:27
DC is like there isn't an obvious like you know you know going to be a path forward over the next week or so and then Speaker Johnson isn't even necessarily like releasing a schedule on when he's going to call everybody back to even you know sort of you know fix it. Um so it's an interesting place
2:03:38
it. Um so it's an interesting place where you know uh CRS and shutdowns generally favor incumbents right um and so but what's interesting though this year is like now a bunch of the net new tech companies are kind of the incumbents right um and so you know if
2:03:56
you were and in 20 whatever 18 government shutdown pretty painful if you're like androll in 2025 yeah it's I mean you'd prefer for you know budgets to keep going but like you can actually probably even continue to beat plan, etc., even with, you know, sort of
2:04:09
, even with, you know, sort of government shutdown.
2:04:10
Um, and so mostly what it makes it painful for is some of the like, you know, nextgen, you know, companies that have been started in like the last year or two, really, you know, sort of painful.
2:04:18
Um, so at least as somebody that's closer to an incumbent now rather than a 2-year-old company, um, I don't mind the government shutdown, you know, sort of that much.
2:04:27
If anything, it kind of gives me like more room to like, you know, get away from some of the, you know, sort of C, series A, series B companies. So, you know, Mr.
2:04:34
Politician, you know, shut down away and, you know, we'll keep flying capsules.
2:04:36
It's also uh you know if the government shuts down then if it restarts it'd be a bullish catalyst >> send us to new all-time highs.
2:04:46
>> Close us out with a white pill.
2:04:46
What's the what's the best new development happening in defense tech space?
2:04:50
Any positive news aside from the government stuff?
2:04:55
The shutdown's kind of disappointing.
2:04:57
>> Will you be buying the Uni Tree IPO? >> That's depressing.
2:05:01
Um, you know, >> no, here here's what's depressing is it's going out at 7 billion, >> what does that say?
2:05:08
If they're the lead, if they're the leader and the and the market says this is a 7 billion company and we have companies that are worth 40 billion.
2:05:16
If you look at Tesla, what what premium does does Optimus give to Tesla?
2:05:21
We're clearly valuing the potential of humanoids more than uh the at least the the Hong Kong stock exchange.
2:05:27
the Hong Kong stock exchange. This is kind of related to like how I think you could use unitry for peace where if I were to provide like you know sort of the analogy what Apple is to Foxcon you ideally need some US company to be to unitry where it's like both the US government and the CCP are not happy
2:05:46
with the Apple Foxcon relationship right um the US government is not happy that Apple does so much of its manufacturing with our primary adversary and would like them to ideally push to relocate more of that to more allied nations like you know Vietnam, India etc which they're starting to do um and do more of that in the United States. The CCP is
2:06:00
The CCP is not particularly happy that Foxcon puts in all this work, technology, process, etc.
2:06:07
, ships out these iPhones, and then that just ships a ton of profit dollars over to the largest company that's owned by their adversary.
2:06:13
And so both sides are super unhappy with the relationship.
2:06:17
And yet the relationship has now persisted for like over 20 plus years, right?
2:06:19
And so I think there's the downside case of Unitry succeeds you know a lot in China and just makes like humanoid like you know soldiers that like end up you know invading Taiwan.
2:06:29
That's like the downside scenario.
2:06:29
The upside scenario is Unitry becomes a contract manufacturer, but like an American company has way better foundation model, AI controls, you know, sort of robot, whatever operating system and design and uses Unree basically as like the manufacturer.
2:06:42
And then you basically get into this like unhappy wedding on both sides in a way that actually makes it even less likely that basically China actually >> only downside is a sci-fi scenario where there's a backdoor into the, you know, millions of humanoids that get deployed into the US.
2:06:57
And >> yeah, >> I mean, >> yeah, but but my iPhone's not going to get up at night and stab me in the chat.
2:07:04
>> I mean, it could explode and it's in your you know, I mean, like, look, like you thought the Hamas like, you know, our pagers were bad.
2:07:09
Let me tell you, there's like a little bomb in each of these and like you can explode like just government officials, you know, >> could be could be tough.
2:07:18
>> Yeah, I asked for a white pill.
2:07:18
That was >> and I got a black pill.
2:07:21
>> Close, but we'll get you next time, Delian.
2:07:23
Thank you so much for hopping on, taking the time. Congratulations.
2:07:26
Congrats on the little one.
2:07:26
We will talk to you soon.
2:07:29
You heard him talk about eight sleep.
2:07:31
You can enjoy an eight sleep. If you go to eight. com, get a pod 5.
2:07:36
A 5year warranty, 30 risk-free trial, free returns, free shipping.
2:07:39
>> How did we do last night? >> I got cooked. I got a 78. >> I got a 76. >> I got a 73. So, you beat me. 73.
2:07:45
Um, our next guest is from Flock Safety, a founders fund portfolio company.
2:07:52
Actually, uh, we have Garrett Langley.
2:07:54
I'm very excited to come in from the Reream Waiting Room. Garrett, how are you?
2:07:58
I'll let you take the intro. >> Welcome to the show. >> Good. >> How are you >> doing? Great.
2:08:03
Uh, been uh we've been super excited to have you on uh first time.
2:08:07
Why don't uh you give a quick introduction for anybody that's been uh living under a rock. >> Yeah.
2:08:14
Or hiding under a rock maybe.
2:08:14
Um but uh yeah, Garrett Langley started a company called Flock Safety with a couple friends eight years ago and we catch bad guys.
2:08:24
Um we will help local law enforcement make just around 700,000 arrests across the country this year.
2:08:29
Um all violent nonviolent crimes.
2:08:35
>> What was the uh what year?
2:08:35
How quickly what was the time to first arrest?
2:08:41
>> Yeah, just under 60 days. So, we were in Y.
2:08:43
Yeah, we were in YC and it was horrible because we had built this product.
2:08:47
We had a couple neighborhoods using it >> and we were like, demay is going to be a bust cuz like yeah, we built a camera, but like what else?
2:08:54
And I kid you not, the week before this guy broke into someone's home, stole a nice road bike, Dicab County, Georgia, made the arrest, and so we went on demo day.
2:09:03
We had one slide and it was this mug shot.
2:09:05
And we were like, we build cameras that do this. Please invest. >> Wow.
2:09:11
We already had the mug shot. Yeah. Yeah.
2:09:13
That that uh better than like, you know, we got 2,000 MR. Yeah.
2:09:19
Real world real world results. Um that's great.
2:09:23
And feels more, you know, more critical than ever with everything going on in the world.
2:09:27
What is What have you guys been up to lately?
2:09:32
>> Yeah, I mean, so we're live in just over like 6,000 cities now.
2:09:36
So pretty well deployed um and building a lot.
2:09:40
So we got into the drone business earlier this year, kind of at the end of last year, earlier this year.
2:09:46
We've now got drones flying all over the country.
2:09:48
Um just kind of launched that also for the private sector.
2:09:52
So it might surprise you that private inter enterprise spends north of $30 billion a year on unarmed guards.
2:10:00
>> I think drones are a pretty good alternative. They're cheaper. They're more reliable.
2:10:03
And so now like if you're >> and and unarmed guards are are effectively just providing like a deterrence, right?
2:10:10
They're not meant to engage, but they're just trying to show that there's that you're not, >> you know, you're not alone basically, right?
2:10:18
And so so drone can can play a similar role and potentially even a better role if it can follow a bad actor, something along those lines. >> Yeah, I know. Exactly.
2:10:27
I mean, you think about like a retail example, which if you're in the Bay Areas happens a lot, stolen car car gets stolen, they drive it to a Home Depot to a Target, we know that car is stolen.
2:10:38
Today, we just notify 911 and, you know, depending on what city you're in, a couple minutes or 20, 30 minutes later, 911 shows up.
2:10:44
Now, with our drone, that drone can get automatically dispatched off a rooftop out of a box, completely remotely, kind of teleaoperated.
2:10:53
You've now got live video streaming to local law enforcement inside their vehicle. This is the vehicle. This is the person.
2:11:00
They're going into the store.
2:11:02
They're leaving the store, whatever it may be.
2:11:03
And so, it's kind of the way to think about it is we just treat drones like a camera that can fly.
2:11:08
So, we're pretty good at building cameras.
2:11:09
And now these cameras can fly and chase bad guys.
2:11:14
>> Can you give me some uh timelines on when uh you know, the average American would be seeing just drones flying around?
2:11:22
what scale we're going to see that.
2:11:24
Are you bullish or excited about humanoid robots or wheeled robots?
2:11:26
We've seen what Starship Technologies has done.
2:11:31
There's a whole bunch of companies that are putting, you know, more like simpler robots on the on the ground.
2:11:38
Like what what's the actual path and and what kind of timelines are you kind of operating against?
2:11:44
>> So, I mean, the the good news is that these drones fly high.
2:11:48
So, we fly 400 feet up in the air. >> Okay.
2:11:50
So, if you're not looking up, you're not going to hear it.
2:11:53
You're not going to see it. >> Wow.
2:11:55
>> Um, the other thing is our drones can cover up to probably 30 30 square miles effectively. >> Yeah.
2:12:03
>> And so, the sheer number of drones you need actually isn't that many. >> Sure.
2:12:07
>> So, if I think about a county near me, it's the largest county in Georgia.
2:12:09
I think it's just over 500 square miles.
2:12:14
They're going to get 20 drones to cover the entire county. >> Wow.
2:12:18
>> And that's a couple million people.
2:12:18
like you're never going to see these things.
2:12:21
>> So, you would probably wind up either partnering or building yourself like something that looks kind of like a cell phone tower, has charging infrastructure, maybe there's a human that goes out there and fixes parts if there's problems.
2:12:32
>> Does it even need to look like that or can you just put it on a roof of like a Walmart >> put >> even better fire department?
2:12:37
Fire departments are legally required to be equideployed in a there's a federal regulation of the density of fire departments which makes them a perfect place to also put drones for sure.
2:12:46
So, you put them on top of a fire department.
2:12:50
>> Uh, they're fully autonomous. They live up there. It's got HVAC.
2:12:51
It's pretty cozy place to live as a drone. Yep.
2:12:55
>> And then someone clicks a button uh from a computer just like you guys are on.
2:12:59
>> Y >> drone flies and then it can track the car, locks in, tracks the car or the human, it's locked in.
2:13:03
You can do multiple drones to one operator.
2:13:05
Um, so even in a county like the one I'm describing with 20 drones, you know, they probably need three or four operators at all times, >> but like they're covering 500 square miles. like insane coverage. >> Yeah.
2:13:17
Talk to me about some of the other robots.
2:13:19
Uh robotic dogs, wheeled four-w wheeled robots, humanoid robots.
2:13:23
Like when does this >> Well, and and before that, like talk about autonomy timelines because if you're flying 400 feet in the air, >> at what point, like there's not a lot else going on up there.
2:13:35
I imagine uh you could build these to be pretty fully autonomous pretty fast.
2:13:40
Yeah, I mean the the the autonomy is already there.
2:13:44
Um the issue is the FAA regulation.
2:13:46
It's like right now we are legally required to have a human click a button that says launch drone. >> Yeah.
2:13:54
>> And then they need to have a part 107 license that says they can fly >> and they're a pilot slice.
2:14:01
Uh >> it's a it's an easier version.
2:14:03
It's maybe 20 or 30 hours of you know studying specifically for commercial drones. >> Yeah. >> Yeah.
2:14:11
And I think it's good, but the autonomy is there.
2:14:13
Um, and the FAA has done a great job over the last few years of kind of like trying to catch up.
2:14:17
They were a little slow for a while. They're catching up now.
2:14:19
I think they're making better decisions.
2:14:21
Like they're now allowing us to fly multiple drones with one operator.
2:14:25
Uh, it used to be it had to be one to one.
2:14:26
It used to be you had to be able to see the drone.
2:14:28
Now you don't have to be able to see the drone.
2:14:31
So in the case of uh that big county, that person sitting downtown, they're flying a drone 50 miles away. >> Yeah.
2:14:39
>> And I think that makes a lot of sense.
2:14:40
So the autonomy is there.
2:14:40
It can autonomously track the car, stick to it so you can avoid a high-speed pursuit.
2:14:45
Uh but the FA is going to make sure that it's safe.
2:14:48
And I agree like safety is the most important feature >> we sell. >> Yeah.
2:14:52
How much of >> it's a big thing.
2:14:53
>> It feels like high-speed chases are like uh definitely a bug, not not a feature, right?
2:14:59
Because it creates an entirely new danger, right, of of hey, let's get we've got one guy that stole a car.
2:15:07
Let's get six cops >> driving super fast. What can go wrong?
2:15:12
>> Yeah, >> it's probably thrilling though.
2:15:15
>> I think it's like probably one of the highlights of the job.
2:15:16
Um, if you're like an adrenaline junkie because like you're driving 90 m hour trying to like chase someone.
2:15:20
No, it's definitely a bug and I think most elected officials would agree as they've started to ban it which creates all types of new conflicts.
2:15:25
So, I think highspeed pursuits are a really good one.
2:15:30
>> I think the other thing that we've been surprised by is the reduction in service calls.
2:15:35
And let me try to expand on that >> please.
2:15:38
>> When you call 911, the average response time might be 30 minutes. >> Mhm.
2:15:42
>> And a part of the problem is it's just slow to drive.
2:15:43
So we'll get there faster because of that >> because we don't have to wait in traffic.
2:15:47
The other issue is that a lot of the times you'll call 911 and say, "Hey, these two guys are in a street fight."
2:15:51
And then 20 minutes later those guys are gone.
2:15:54
We still had to send an officer.
2:15:57
They still had to look around.
2:15:59
Maybe they go into the gas station and grab a Gatorade because they're thirsty.
2:16:02
and all of a sudden like we've wasted hours of this officer's day.
2:16:03
So, one of the beautiful parts of the drone is actually it gets there faster and it can also remove that call for service from the backlog so that for actual critical incidents, we get humans there faster as well.
2:16:17
I think the other one that I'm pretty excited about is we have a number of colleges that we work with that are using drones to help escort females at night back home.
2:16:26
>> So, you can call 911 and be like, "Hey, I don't feel safe. Like, it's really dark. I just left this party."
2:16:30
And you don't feel great, I think, as like a 19-year-old maybe doing that because, you know, now I'm gonna get in the back of a patrol car. >> Yeah.
2:16:37
>> Now, in a couple colleges we work with, they're pushing this out like, "Hey, you can always call 911 and a police drone will escort you home." >> Wow.
2:16:44
>> And we'll have video, we'll have eyes on you so that if anything starts to look suspicious, like we're on top of it.
2:16:47
And I think there's just a ton of use cases that we're just undercovering now um to make drones just a more bigger part of our daily life.
2:16:55
If you run the the like economic calculation on the it's 30 square miles for one drone and there's cost to service that and batteries versus do some sort of deal where every stoplight has a camera on it and you have sort of coverage 360 cameras on every single corner.
2:17:13
Uh is there an economic calculation here that makes sense where the drone's more efficient or is it just more flexible?
2:17:20
Like what are all the trade-offs that a city or you think of? Yeah.
2:17:23
So the the way we think about it is um there's a certain cost per citizen >> that a city is comfortable paying to eliminate crime.
2:17:33
>> And we do think you can eliminate it.
2:17:34
Like you can't eliminate uh emotional crime like crimes of hate, crimes of passion.
2:17:39
>> But the capitalistic crime like there was a really funny no jumper uh podcast a couple weeks ago. Yeah.
2:17:45
>> Where the guy was like those effing flockers >> uh they're making it too hard to commit crime in San Francisco.
2:17:50
And I'm like that that that's the goal cuz like that guy knows >> on a podcast on a podcast. >> Wow. He called you flockers. That's hilarious. >> It's brutal. >> Yeah.
2:18:00
It's just about like it's about making it so that >> somebody could still do it but it's not economically viable. >> Random podcast. That's amazing. Yes.
2:18:10
>> That's product market fit.
2:18:10
Getting called out on no jumper for for stopping crime. >> For stopping crime.
2:18:14
So when we look at that though and like you know the the the probably the most deployed city we have is spending about $22 a citizen per year with flock.
2:18:23
I think that's pretty reasonable.
2:18:26
That's not actually not that much money.
2:18:28
Um it's a lot for that that city but I mean as a individual I believe that city is going to solve every single crime that happens.
2:18:33
I don't think that's too much of a burden.
2:18:35
But I do think it's about you you think about building a cake.
2:18:39
You got to have different layers.
2:18:40
You don't want just all cake all icing. You want cameras.
2:18:44
You want cameras that fly. You also need software.
2:18:46
Uh you need trailers because there's a more of a deterrence.
2:18:47
So I don't think it's a one-size fit all fits all.
2:18:51
But for us right now, we kind of see that deployments where you're going to cover your perimeter with cameras that track cars.
2:18:56
You're going to have kind of PTZs, panel zooms for your major intersections, and they're going to use drones to kind of cover the whole city or county as an overlay. >> How How do you Sorry. >> Yeah, I know.
2:19:09
We're going to get there. Yeah.
2:19:11
I really want to know because Prosper Citizen is probably really expensive.
2:19:14
We're seeing stuff from Unitry.
2:19:14
There's companies in America. Optimus is far away.
2:19:16
I just give me your human take. Sorry.
2:19:20
>> So, so my so my humano take is like as it stands today, they're just way too expensive.
2:19:24
>> Um, so if you think about like an average retailer, some of them make a lot of money, right?
2:19:29
If you think about like a big box retailer, they might they might generate a couple million in eBay per location, but then you go to some do you kind of go down market to, you know, a um an Ulta, Sephora, a Dollar General.
2:19:44
Their box profit is just like not very high.
2:19:46
They have a lot of locations. >> Yeah.
2:19:48
>> But their individual stores are actually not very profitable.
2:19:50
They can't spend more than a couple thousand dollars a year per store in safety.
2:19:53
I just I don't see a world where other humanoids or or dogs worked there.
2:20:00
And we're seeing that unfold as well where bigger footprint locations just have both more assets and more dollars.
2:20:08
In my conversations with retailers in particular, the kind of two- wheeled approaches have been laughed off.
2:20:16
>> They get stuck in a corner. >> Yeah. >> They get kicked over.
2:20:18
The dogs are seeing more efficacy.
2:20:22
>> And I got to imagine like a biped like humanoid is going to be the most effective because that just sounds really scary. >> Yeah.
2:20:29
>> Well, now and some of you can't kick you can't kick them over. They come right back up. It's it's crazy.
2:20:35
>> Talk about a scary situation.
2:20:35
You're trying to rob a place, you try to kick it over and it just stands back up.
2:20:38
It's just looking at you just like, "What next?"
2:20:40
>> There's going to be like dark webuto down. >> Yeah.
2:20:44
Dark web tutorials of like how to get it. Yeah.
2:20:46
Like >> all of a sudden the the the bippers bring like >> it's so sci-fi >> like you have to cut these cables on the back and >> it's going to be wild. Sorry Jordy.
2:20:56
Uh you had another question. >> Yeah.
2:20:59
I was just curious about uh just like how you think about product prioritization because I imagine when you get in working with these cities and police departments and they start using your products and getting value out of them, they just start coming to you with like more problems both like hardware software.
2:21:16
But it feels like the drone opportunity is big enough that at some you know it's it's probably hard to >> that you have too many opportunities uh than you could probably pursue all at once. >> Yeah.
2:21:28
Yeah, I mean I think we took the company from one skew to eight SKs in the last 12 months that all have clear of sightes and nine figures of AR. >> Great. >> Yeah, I agree.
2:21:39
Uh and the pipeline Yeah, the pipeline is there. The AR is growth. Like it's kind of wild.
2:21:45
This will be our first quarter, I think, where uh like our our core business like won't be the biggest product line, which is kind of weird.
2:21:52
I think that's a good thing >> um for us because we're not decelerating on that side.
2:21:56
just other products are really attractive.
2:21:59
>> But I think it is, you're right, like uh our bigger concern is the market adoption, not our ability to build.
2:22:06
>> I mean, this is a kind of enterprise type buyer where they're just only going to adopt things so quickly.
2:22:09
Um, so even like on the AI front, we have a lot of ideas.
2:22:13
We've built a lot of interesting products.
2:22:15
They're just they're really nervous to adopt too quickly.
2:22:16
Um, and I think the example I always give is maybe you're trying to book a hotel and you you call the hotel and you realize you're talking to a an AI agent, but you're kind of like, I don't know, this is better than the alternative because I'm just trying to book a hotel. >> I don't know, man.
2:22:30
I think when you like call 911 and and you're in the middle of a violent situation, I think there's something warming about knowing there's another human on their side trying to help you. >> Yeah.
2:22:39
Not to say they can't be augmented with AI, but I think that's like a much more challenging societal question >> of what do we want humans to do and what do we want AI to do? >> Yeah.
2:22:50
>> And we are being conscious to not make that decision for the hundreds of millions of Americans we help keep safe. >> Yeah.
2:22:56
Just as a taxpayer, I would imagine that I would want the person to pick up, but I'd love for the calls to be transcribed and then AI to find patterns between what's happening and insights and analytics and all. >> Yeah.
2:23:07
There's also something strange.
2:23:07
Uh >> yeah, >> like if if you if somebody's calling 911 to report like a drunk somebody that's obviously drunk driving, it's like you're taking up bandwidth from somebody that is a human that could be going elsewhere.
2:23:20
elsewhere. It's like for you know maybe it's like you should there should be no dial time is a clanker immediately picks up and says what's the problem and then routes it either to a human or if it's less like you know if it's not like immediate you know some violent act is
2:23:37
taking place it can actually be solved by >> a voice you know voice model >> and I answer that approach but yeah I think to answer your question like we're pretty focused on doing three things for our customers solving more crime resurrection Now there's about a 40% clearance rate for for killing someone. So like you got almost 50 chance of
2:23:53
So like you got almost 50 chance of getting away with homicide. >> Wow. >> That should be zero.
2:23:56
You should every single person that kills someone should go to jail.
2:24:00
>> Nonviolent crime is even worse.
2:24:00
We want to solve cases faster.
2:24:01
It takes way too long to solve crime.
2:24:04
>> And we want to do more with less.
2:24:04
Like most people I don't know if you all remember our grandparents, our parents, it was an admirable job to go into policing to fire. >> Yeah.
2:24:14
>> This generation does not view it that way.
2:24:15
which means 80 plus percent of police departments are underst staffed against their budget and it's not getting better and technology's going to fill the gap. >> Yeah.
2:24:23
Um people have compared you to like the Anderol or Palunteer kind of like I don't know American dynamism companies or whatever.
2:24:31
>> People have compared you to good businesses.
2:24:32
businesses. Yeah, sounds >> but uh but I'm interested about uh we talked to Sham Sankar CTO of Palunteer about some of the some of the difficulties of deploying uh software into governments and he gave the example that one time they they built a
2:24:50
beautiful piece of software exquisite system and they went to deploy it and the the the the team on the other side of the government that was firing it up was trying to run it on a computer that had like you know 128 megs of RAM instead of like a gig. And so that that
2:25:02
And so that that was kind of the dawn of the forward deployed engineer.
2:25:06
And I'm wondering if you uh if you have any unique solves on how to deploy systems, how to deploy software, if you like the forward deployed model or something else, like what's working for you at actually getting your solutions into the field.
2:25:21
>> So really similar example, I remember distinctly we were like launching this product and the designer has a beautiful, you know, MacBook and this HD retina display. It's beautiful, right?
2:25:32
super expensive setup and she's designed this like super slick thing >> and I was like great Mariana like you should go like field test this like this looks dope like let's launch it. >> Yeah.
2:25:42
>> And she gets into a patrol car and it's a Panasonic Tough Book. >> It's a 13inch screen.
2:25:46
It's 1280 by 1024 resolution and he's using it while he's driving 40 miles an hour.
2:25:54
>> And I'm like this the product's not going to work. Like we got to go back.
2:25:58
That that is literally our we've got I don't know hundreds of thousands of DAUs using this product now and they're very happy.
2:26:03
But that's their normal use case. They're driving. >> Yep.
2:26:06
>> They're on a tough book. Like it's a tiny screen. >> Yep. >> It's low res. It's old.
2:26:12
>> And so like you're pretty much treating it almost more like a mobile app but like with a distracted teenager as your customer.
2:26:18
>> Uh and that's been pretty difficult because we want to build these like super high fidelity.
2:26:20
You're just sitting at a desk all day doing your job and like our customers are in the field all day.
2:26:25
Um, so that one's been pretty pretty tough.
2:26:28
And then I also think the other burden we've done a good job of going around is just it in general. >> Yeah.
2:26:34
>> Like I city IT is really tough to work with.
2:26:37
They're really tough to kind of just like get get going.
2:26:38
So we tend to build everything we can to avoid it.
2:26:43
>> There's some pros and cons to that, but um, so far I think it's been mostly pros.
2:26:47
>> How does that actually work? I imagine.
2:26:49
Is that just like you're only looking for like user licensing or like OOTH or something like how at some point you have to plug into the IT systems a little bit I imagine like >> well is it so as a good example um historically everything in our world is on prim >> like we're one of the first cloudon solutions in this market. >> Yeah.
2:27:12
>> And early on that was painful because people are used to buying a piece of software they install a desktop application that installation requires IT support.
2:27:18
And so building everything in the cloud, which for this, you know, for y'all it's like, well, obviously that's a very it was a very contrarian take seven years ago and now it's moving.
2:27:28
But I mean, we have a customer >> that just a few years ago >> moved off of paper records and they had filing cabinets.
2:27:35
Like imagine if you report a crime, they're like, "Hold on, we got to go find that file." >> Wow.
2:27:39
>> And that this is a big city.
2:27:39
This is city with millions of people in it. >> Yeah.
2:27:42
>> Um just got off of paper.
2:27:42
And I think in Maryland in 2023 the cloud became legal. >> Mhm. >> The cloud is people.
2:27:54
>> This is we're going to protest for legalizing the cloud. We love the cloud.
2:27:58
>> If you need if you need if there's any cities that have still banned cloud, we will >> and we and we hate paper.
2:28:01
We are we are the show is presented by RAMP of course and we are strong in favor of ending paper and going to the cloud. >> Agree.
2:28:10
But literally, we didn't do business in Maryland until 2023 because it was illegal to be in the cloud. >> Yeah.
2:28:15
Did you ever get a a a test environment with a tough book and a car that a designer could actually go and and and drive around the parking lot of your office or anything like that?
2:28:27
>> No, we make all of our employees go do ride alongs. >> Okay. Yeah.
2:28:30
>> It's the It's so much fun. >> Yeah.
2:28:31
Because you get to see the product in action.
2:28:33
You get to meet the customer.
2:28:35
You get to go catch some bad guys.
2:28:35
Like it's a pretty fun way to spend eight hours. >> Yeah.
2:28:38
uh h how are you thinking about data privacy where data lives?
2:28:41
There's obviously so many advantages to being cloud-based, but then you have some citizens who might say, "Wait, what?
2:28:51
Like, why is my image with this private company? I I'm a taxpayer.
2:28:53
I I I don't have a vote at your board meeting.
2:28:55
I have a vote at my city council."
2:28:57
H how do you grapple with all that?
2:28:59
What's what's the mood on the ground in various cities?
2:29:03
What are the dividing lines?
2:29:04
>> Yeah, I'm glad I'm glad you mentioned cities.
2:29:06
So our general stance is we don't write the rules, we create the levers for local politics to dictate.
2:29:13
>> And so data retention is the easiest one.
2:29:16
>> In some cities we work with data retention is seven days.
2:29:19
>> And every single data we capture is purged after seven days.
2:29:22
>> But then in some cities like Dallas, it's a year.
2:29:24
In New Jersey, it's five years.
2:29:26
A state legislation for five years. >> Wow. >> I don't care. I live in Atlanta. Yeah.
2:29:31
>> I think we're like at 30 or 60 days here in Atlanta.
2:29:33
And that's up to the local politicians to decide like what makes sense for them.
2:29:38
>> The other thing that we do that's pretty unique is every single action in our system is audited in perpetuity.
2:29:44
>> So whether you're downloading, you're searching, you're doing something that's stored.
2:29:47
Y >> so that the city manager, the internal affairs bureau, those groups can actually say, "Hey, has anything nefarious been done?" >> Yeah.
2:29:54
>> Um that's that's a big concern of people.
2:29:55
So those two things tend to be a pretty pretty good levers for letting local politics dictate versus to your point, no one elected me the police chief of America.
2:30:03
I just built a camera at my house.
2:30:07
>> I want to pitch you >> a startup idea that Blake Scho, the founder of Boom Supersonic, he was thinking about doing this.
2:30:15
Have you heard this story?
2:30:16
He was thinking about building something before he started Boom.
2:30:19
Probably right around the time you were starting Flock.
2:30:21
He wanted to make a smart uh stoplight that would have a camera on it and it would see, hey, there's no cars here.
2:30:29
Why do I have a red light? Uh switch them.
2:30:31
Feels like it would improve traffic flow.
2:30:33
Uh it's a it's a you know, a situation we've all been in.
2:30:38
You're sitting there at the red light.
2:30:39
There's no one coming the other way. Flip the switch.
2:30:40
Uh why is that is that a good idea?
2:30:44
Is that something you could do?
2:30:45
Is that something any company could do?
2:30:47
Like what's the market structure of the stoplight industry?
2:30:48
Do you need to do some sort of private equity rollup?
2:30:52
Talk me through your your if you were putting on a VC, how how would you interpret that pitch?
2:30:57
>> My my my push back on that pitch is to show me the incentive and I'll show you the behavior.
2:31:02
>> Who Who's getting promoted?
2:31:02
Who's making more money if we fix traffic? >> I don't think anyone.
2:31:10
>> The opportunity cost of the workers, it's all diffuse in the economy.
2:31:14
>> No, but this is the but don't you think don't you think a mayor could run on, hey, we have a lot of congestion here.
2:31:21
We have a techn You'd have to do a study or something to show that that you could inc that reduce congestion if you had smarter traffic lights.
2:31:27
Like feels like some >> mayors don't mayors get elected on two things. >> Yeah.
2:31:34
>> Well paved roads and low crime. >> That's it.
2:31:37
>> It's the only things that matter.
2:31:39
>> Like so I mean I think the product idea makes a ton of sense. >> Yeah.
2:31:43
>> But I just go back to like one of the only reasons why because you mentioned Andre. Yeah.
2:31:47
When we were pivoting the business into local government, Trey Stevens was like, "Do not go sell to police. It's a horrible market.
2:31:55
>> We tried it at Palanteer. It's not gonna work. >> It's not gonna work."
2:31:58
I was like, "Well, I know like I'm a little bullish." He was an investor.
2:32:01
He's an investor in the company. I think this might work.
2:32:04
>> Um, and what I didn't realize is >> systems like Palunteer are too many steps removed away from solving crime >> at a local level.
2:32:13
And like Flock had coincidentally built a product that solved crime right away.
2:32:19
>> And so for a police department, crime is equivalent to revenue.
2:32:21
So we weren't helping them save money.
2:32:22
We're helping them generate revenue.
2:32:23
And like that does get a sergeant, lieutenant, a police chief promoted.
2:32:27
That gets a mayor reelected.
2:32:29
And and that's that that would be my pitch to Blake is like we got to find a way to make this relevant to a mayor. >> Yeah.
2:32:36
>> And like just a little bit less traffic isn't going to move the needle. >> Yeah.
2:32:39
>> And have them move off from having no technology and no cost structure.
2:32:41
Well, well, he's he's busy building supersonic jets.
2:32:45
He's got a good thing going for now, but maybe maybe that's the next thing he works on.
2:32:48
Uh I have a second startup to pitch you based on your answer to that question.
2:32:52
Uh automated road paving startup.
2:32:55
I go to the mayors and I say road paving gets you paid and gets you in the job like would that work? Is that a good idea?
2:33:02
How would that play out if you you know thought that process through?
2:33:05
What what advice would you give to a to a road paving >> entrepreneur?
2:33:08
Yeah, there was a YC company doing the little road >> I remember this. Yeah, I remember this.
2:33:12
But yeah, yeah, yeah, walk me through your your thought process on this.
2:33:15
>> So, I actually think that's a great idea.
2:33:16
Um, we're working on something >> tangential to that.
2:33:21
>> Um, so one of the biggest issues is that to do what you're describing, you have to know where the potholes are.
2:33:27
>> And cities have no idea where potholes are. >> Yeah.
2:33:29
>> And they rely on you calling 311, which I'm sure you both have done many times and said, "Hey, I'm just calling in to report a pothole." >> Yep. >> No, you don't.
2:33:35
And that that also costs the city on average like $8 per call on a 311.
2:33:40
>> So, it's a really inefficient system. Okay.
2:33:41
>> And so what we've been able to do is train our cameras to look for potholes.
2:33:46
>> And so now we can actually report back general road uh road conditions. >> Yep. >> Cool.
2:33:52
>> And that is the beginning of them able to say, "Hey, now we can actually be intelligent about where we should send people to go repave." >> Yep.
2:33:57
>> Uh and I think an automated robot that does the paving would be even better.
2:34:01
But yeah, I love that idea.
2:34:01
I think that I'm I'm in $25,000 check. >> Fantastic.
2:34:06
Well, you know where to reach him. He is of course.
2:34:10
>> Oh, last uh last did you uh wanted your reactions.
2:34:13
We had a guy named Riley on yesterday who made a find my parking cops. Did you see that?
2:34:19
>> Oh, that that park was awesome. >> Yeah. >> Yeah.
2:34:22
>> He's a little shut down back up.
2:34:23
>> It got shut down very quickly.
2:34:23
I was impressed with San Francisco's reaction time on that.
2:34:28
They shut it down like very quickly.
2:34:30
But, uh >> it's good fun.
2:34:33
>> Yeah, >> I'm a fan of it.
2:34:34
I think most people don't pay for parking because most of the time you don't get a ticket. So >> yeah.
2:34:39
Well, uh, thank you so much for hopping on the show.
2:34:41
This >> great to get the update. >> Yeah.
2:34:44
And congratulations on the progress and thank you for everything that you do. >> Come back on anytime.
2:34:48
>> Yeah, we'd love to talk to you soon.
2:34:50
>> Have a great rest of your day.
2:34:52
>> Uh, let me tell you 10 9 figure businesses. >> Yeah.
2:34:55
If you want to reach people in cities, you got to go to adquick. com.
2:34:59
Out ofome advertising made easy and measurable.
2:35:00
Say goodbye to the headaches of out of home advertising.
2:35:02
Only AdQuick combines technology.
2:35:04
out of home expertise and data to enable efficient seamless ad buying across the globe.
2:35:08
>> Before uh our next guest joins, we did need to acknowledge that Mark Leonard has stepped down from constellation software.
2:35:15
Sad news for health reasons. >> We hope he's okay. Uh interesting timing.
2:35:20
Basically called the top on >> Yeah, it was uh we talked to Carrie No interest on the show.
2:35:25
You can go pull that up if you want to hear our conversation with him about these private equity software enterprise software rollups.
2:35:32
Uh Mark Leonard was mixed on AI un sort of unclear on on how it would affect the business long term but the market did not like it and the stock fell off and the stock has fallen again on this uh very sad news.
2:35:44
So we are of course sending him our best.
2:35:46
The health reasons that he is stepping down for are unspecified at this time.
2:35:50
So uh but please send your thoughts and prayers to >> thoughts and prayers >> uh Mark Leonard.
2:35:55
And we have our next guest in the room waiting room TVPN Ultra Dome.
2:36:00
And there he is suited up.
2:36:03
>> Matan, how are you doing? >> Hello. >> I'm doing well. How are you guys doing? >> It's been too long.
2:36:07
Yeah, we've seen you before in the TV panel. Welcome back. >> Pleasure. >> Give us the news.
2:36:12
Give us the give us the 140 characters and the company.
2:36:16
Refresh everyone and then give us the news. >> Absolutely.
2:36:19
So, I'm Maton, CEO at Factory.
2:36:21
At Factory, our mission is to bring autonomy to software engineering.
2:36:26
what that means more concretely.
2:36:26
We have built droids which are autonomous software development agents and more importantly I'm here to tell you that they are the number one agent in the world.
2:36:38
Um >> as we just released today in the terminal bench benchmark which evaluates tools like claude code openis's codec cli factory is number one >> and also number three and also number five. Pick any model you want.
2:36:51
Factory is still number one.
2:36:53
No matter how you slice it, um, our droids are simply the best agents in the world.
2:36:58
So, excited to share that. >> Yeah. What's the secret? What what got you there?
2:37:03
Is it uh pre-training, post- training, a bunch of RL stuff? Is it a data flywheel?
2:37:07
Like what's what's driving the growth? >> Yeah, great question.
2:37:11
So, I think one of the biggest things is that most if not all of the agents out there are built for one model in particular.
2:37:16
So, you know, Claude Code is built to work with the Claude models.
2:37:20
OpenAI's codeex is built for the GPT models.
2:37:23
There are some uh other tools out there that out there that aren't from the research labs that are focused on really like fine-tuning their agent harness for one given model for any given step.
2:37:34
But what we've done with our droids is make them fully model agnostic, which actually makes them more performant in the long run.
2:37:40
It's kind of similar to if you were a human engineer and you only let's say studied one coding language.
2:37:46
You would actually be a weaker engineer than if you studied all of them.
2:37:50
It's somewhat analogous there in terms of how we've built these droids that now allows them to be the most performant with any model that you that you put in under the hood.
2:37:59
>> How are you viewing the market and the customer landscape?
2:38:01
uh automated software engineering that feels niched down from years ago which was just uh the the the transformer-based LLMs can write code.
2:38:12
Uh but as we actually dig in we see that the needs of a Fortune 100 customer are maybe different than a mid-market company which is different than a startup which is different than an SMB um which is different than a solo vibe coder solreneur.
2:38:26
Uh where are you seeing opportunity in the market?
2:38:29
How bifurcated is the market?
2:38:31
It seems like you've wanted to be very general on the model side.
2:38:34
Are you also trying to be general on the customer size side? >> Yeah, great question.
2:38:38
I mean, I think the the thing that a lot of the other tools have missed is the fact that the further you go into the large enterprise, the less overlap there is between software development and coding.
2:38:50
So, for example, if you're a solo developer, >> your software development is basically just coding.
2:38:54
You're not really doing code review.
2:38:56
you're not doing that much in the way of documentation or testing or uh design docs.
2:38:59
And so where we focus and why we call droids software development agents as opposed to coding agents is that we focus on that whole endto-end software development life cycle which is really where a lot of enterprises are missing.
2:39:13
Like there's study there's the famous study that came out from MIT that said something along the lines of 95% of AI adoption efforts are failing in the enterprise. >> I remember. >> Yeah. Exactly.
2:39:24
And so I mean it kind of makes sense like if you think about software development as as a pipeline.
2:39:30
>> Sorry I love hitting you with the surprise sounding.
2:39:32
We'll get some more positive.
2:39:35
Uh >> we're not bearish on you. It's just that was good. It was perfect. It was perfect. It was just too good.
2:39:39
Um but you know software development is a pipeline right?
2:39:42
And if you focus on only one part of that pipeline namely code generation. Yeah. >> And you expand that.
2:39:49
>> I used to be a physicist.
2:39:49
I don't know if you guys are big on fluid mechanics, but opening one part of a pipe and doing nothing to the others, you don't actually increase the throughput.
2:39:55
Sure, you just create new bottlenecks.
2:39:57
And so, the reason why a lot of these efforts are failing for software development is they're focusing on just coding, but then that just punts the problem down to testing or code review or documentation.
2:40:10
Um, and that's why a lot of the other tools out there aren't seeing kind of the adoption and the real business outcomes that we are with uh the Droids.
2:40:18
How do you see coding agents fitting into the consumer world?
2:40:24
I this might not be your business at all, but I've noticed that the you know I've seen incredible results just from asking a question that you could hit deep research with, but if I ask it to claude code or codecs, it can build an entire HTML web page with JavaScript widgets and bar charts and it just produces it instantiates the information in a in a much cooler.
2:40:46
It's just interesting way.
2:40:48
And I'm wondering how you're seeing uh crossover from advances in automated AI coding into uh a consumer world where they might not even know that they asked for code, but they got code.
2:41:02
It's already happening a little bit.
2:41:04
When you hit 03 Pro, it'll write some code to give you the answer.
2:41:08
You don't even see it unless you unfurl the the reasoning tokens.
2:41:10
But uh how do you see that playing out in like kind of the mid to long term? >> Great question.
2:41:16
I think this is something that uh people aren't yet fully aware of but this is actually the very first statement of our announcement which is the best agents for software development are becoming the best agents for everything.
2:41:29
>> And the reality is because basically every problem can be broken down into some sort of software development problem.
2:41:36
Um and you can actually take it from uh Anthropic themselves.
2:41:39
Uh literally yesterday, Alex Albert, the guy uh who runs Devril at Anthropic, tweets out, you know, the best coding model will be the best model for many types of knowledge work.
2:41:50
Code is how computers operate.
2:41:52
Anything you do on a computer can be done through code.
2:41:54
And this is really why factory being the number one software development agent is really going to then lead to a lot of other tasks.
2:42:03
And we're already seeing this in the enterprise.
2:42:04
So we sell explicitly to developers, but already PMs, designers, even people on the operations side, so like finance, uh, bisops, that sort of thing.
2:42:14
>> We find that they're like sneaking their way in even though they didn't actually, you know, put any budget towards it.
2:42:18
And they're starting to use droids kind of, uh, behind the back of some of the VPs of engineering.
2:42:23
And this is really something that we've expected because of this fact that that code is really the language that computers speak.
2:42:29
And if you want a computer to do something for you, you will eventually through some means need code. >> Uh couple questions.
2:42:36
What what models are you guys getting the most leverage out of today? Is it a mix?
2:42:42
Are you are you focused on you said earlier I think you're focused on a variety being trying to be model agnostic?
2:42:49
But uh be curious where who uh who you're paying out on the back end. >> Yeah.
2:42:56
So right now we achieved the number one score across the board on terminal bench with um Claude Opus.
2:42:59
Um but again we are you know really focused on having the the model agnostic stance because org by org they might have different preferences also what the best model is depends you know what day it is you know tomorrow there could be another new best um and it's important that we have that ability to quickly swap them in.
2:43:22
Um, so that's one thing that I think is important.
2:43:24
It's also even task by task.
2:43:25
There are certain models that are better and we want to make sure that the user has that familiarity uh with the model so that they can go in and, you know, swap it out if there's a task that they're doing that GPT5 might be better for or Grock might be better for.
2:43:38
>> Uh, you forgot to mention you raised some new money today. >> We got a gong here. We got a gong here. >> That is right.
2:43:44
Uh, we have raised >> $50 million from NEA JP Morgan and Nvidia. >> JP Morgan.
2:43:53
>> Jamie Diamonds getting in.
2:43:55
>> He couldn't He couldn't help himself.
2:43:56
>> He couldn't help himself. Couldn't help himself. Couldn't help himself. >> I love it. Congratulations. >> That's our deposits.
2:44:01
Everybody, the world's deposits at work.
2:44:05
>> Uh, talk to me about that last question.
2:44:07
We'll let you get back to your busy day.
2:44:09
Uh, talk to me about the branding.
2:44:09
It feels like uh we just got the idea of an agent recently.
2:44:15
Now you're kind of pitching droids.
2:44:18
It's clearly your term for >> droids you're looking for. >> Yeah. Yeah.
2:44:24
>> People always talk about, you know, being we're not the droids.
2:44:27
>> How much of that is like a differentiated brand that you want to be specific to your company versus like a new coinage that you want to describe a different way of working that you'd actually like to see other companies adopt.
2:44:39
And if you were seeing the entire industry standardized around that particular piece of language, you would be happy or would you be like, "Hey, they stole our brand." >> Yeah.
2:44:49
I mean, I think generally like agent as a term is here to stay.
2:44:51
But the point is as a kind of general purpose term, it's often synonymous with like poor quality or like a while loop wrapped around LLM calls.
2:45:00
And I think what we want to do is to make good on our promise to customers, which is giving them that best agentic experience.
2:45:08
And we do that through droids.
2:45:10
At the end of the day, when you have a cold and you're, you know, your nose is running, you don't want a tissue. You want a Kleenex.
2:45:16
>> Similarly with software development, you want the droids to come in there, right? >> Okay. I love it. I love it.
2:45:19
Well, congratulations on the round.
2:45:21
Thank you so much for hopping by. We'll talk.
2:45:25
>> Thank you guys for having me. Thanks. We'll talk to you soon.
2:45:26
Um, let me tell you about bezel. Get bezel. com.
2:45:28
Your bezel concier is available now to source you any watch on the planet. Seriously, any watch.
2:45:34
And we have our next guest in waiting.
2:45:37
Can't wait to talk about this. Two new launches. One from Meta.
2:45:39
Uh the Alexander announced Vibes, which is a meta AI app for short form AI generated videos. >> Okay.
2:45:48
Yeah, YouTube launched something with uh AI generated shorts leveraging V3 recently. What else?
2:45:54
>> And then uh there's a new chat GPT product called Pulse.
2:45:57
But >> Pulse, we will dig into that.
2:46:00
>> Get into it with our Francis over at Invisible. >> Welcome to the TVO.
2:46:04
Francis, how are you doing? What's happening? Welcome to the show. >> Doing great. Excited to be here.
2:46:10
>> Thanks so much for hopping on the show.
2:46:11
Uh, kick us off with an introduction on you, the company.
2:46:13
Any news you got to share?
2:46:15
What's new in your >> Yeah.
2:46:16
Well, as you saw in Bloomberg, we've raised $100 million this year, which is a big turn of events. >> Go valuation. >> Yeah.
2:46:28
Crazy thing that, you know, when you saw Meta acquiring scale, um, they had raised $1.
2:46:32
8 a billion dollars and we had only deployed 6 million.
2:46:35
We scaled profitably to 134 million of revenue last year. >> That is wild.
2:46:40
I remember I heard about you guys.
2:46:43
I forget what context I heard about you guys for the first time.
2:46:45
But uh even this was probably two years ago.
2:46:50
Even back then I was like you guys were somewhat under the radar.
2:46:52
I felt like but I heard some of your revenue numbers back then. I was like what?
2:46:56
How have I not heard about this company?
2:46:57
But yeah, we wanted to um >> invisible for not much longer. >> Yeah. no longer visible.
2:47:02
Now we're visible by the new domain.
2:47:04
You're visible >> visible technologies.
2:47:09
>> Give us give us an an I mean it'd be helpful to hear like where where you guys have have an edge in in the market broadly and then I I yeah I want to kind of um yeah understand kind >> yeah problem set core uh like what you're actually replacing where customers are getting the most value anything like that would be super helpful.
2:47:30
So 10 years ago when we founded the company, there was a question which is if there's an app for everything, why isn't everything perfect yet?
2:47:36
And Salesforce was the first SAS company in 1999.
2:47:40
For the last 25 years, every enterprise software company has been a SAS company.
2:47:44
And this has actually put customers in a pretty absurd situation where if the customer wants a cake, Silicon Valley will not sell them a cake.
2:47:52
They will sell them tools to make a cake.
2:47:54
And the customer then has to hire a systems integrator like Accenture to stitch together all these SAS applications into an endto-end solution that actually works for the business.
2:48:05
And so uh that is what we were disrupting and Palunteer was not public yet.
2:48:09
So they weren't a well understood comp.
2:48:11
We were kind of like an ugly duckling because we were talking about AI services and um and so it's just a fundamentally disruptive business model.
2:48:18
So you were you were competing uh initially with someone like an Accenture to integrate existing software systems. Is that right? >> Yeah.
2:48:27
We build custom AI applications for enterprises and governments.
2:48:29
And so you can think of this very different than SAS uh like a triangle.
2:48:33
We have the horizontal platform which is very powerful but you have field engineers and field CTOs in a forward deployed motion that builds custom AI applications that actually solve the problem for the enterprise and the government.
2:48:45
And our insight was that even though every you know solution would be custom, the horizontal platform would allow you to sort of build infinitely customizable software over time. >> That makes sense.
2:48:55
Uh we've been seeing Mark Beni off and Dr.
2:48:57
Karp uh going at it over various contracts.
2:49:00
Salesforce which you already mentioned, Palanteer, which you also mentioned.
2:49:04
Um are are you are you in that knockout dragout fight?
2:49:10
Have you found a differentiator around industry or size of problem or or are you just going in the the arena?
2:49:20
>> So Salesforce was founded in 1999.
2:49:22
Palanteer was founded in 2003.
2:49:22
So way before the Genai era.
2:49:24
So we're in a sense like the Gen AI native uh way of doing this.
2:49:28
We hired Matt Fitzpatrick who previously ran McKenzie's Quantum Black Labs.
2:49:32
Black Labs. and uh he's now our CEO and uh after years of being capital efficient this round sort of enables us to go into preipo motion and uh yeah if you're a public markets investor you don't really have that many options for buying AI today uh so and this is
2:49:48
generally true for the enterprises the customers themselves if you want to build custom AI you know solutions uh Palunteer is like a data integration and decision-m support company so they really they really focus on that but if you have a problem in anywhere else in your business. How do you build a custom
2:50:03
How do you build a custom you know solution uh with worldclass field engineers?
2:50:07
Um it's a very different type of engineer, very different type of go to market person, very different type of business that is familiar building custom solutions on the enterprise.
2:50:17
Um you know enterprises do not want their data to leave their systems.
2:50:21
So they they need their uh they need all the solutions to be onrem and containerized and we have the ability to build uh in that way.
2:50:31
uh do you find do do you think that there's a shift in the AI era to large enterprises doing more of these sort of like semicustom solutions?
2:50:38
Is that's going to be a continuing trend?
2:50:42
Because it feels like the last era was very much like rip out the old custom onrem solutions and go to some sort of like one-sizefits-all cloud solution. >> Exactly. Yes. And this is a reversal.
2:50:57
We're going all the way the other way.
2:50:59
Um, and I think it's because enterprises ultimately need proprietary and custom solutions.
2:51:02
So they in the SAS era, like I said, they were stitching together things and they had to do that with some combination of like their internal engineering function and systems integrators.
2:51:12
And systems integrators are not tech companies.
2:51:14
They're going to overcharge, underdel, uh, you're going to pay them by the hour.
2:51:17
So their incentive is to bill you as many hours as possible without getting fired.
2:51:20
Uh, and that's not really aligned with efficiency.
2:51:23
And so um in in this era I think you know enterprises are realizing it's going to be very very difficult for them to build world-class engineering teams and to build a platform like the one we have where uh we can take the best models latest models which we train we have an AI training business um and deploy them in the enterprise uh to actually solve the business use cases that they that they have. >> Yeah.
2:51:48
So so h how much should we read into you guys uh uh transitioning away from training?
2:51:53
Is that a business that is >> we're not transitioning away.
2:51:55
We are fully committed to training.
2:51:56
So actually the the move of meta acquiring scale um basically uh cleared the field and so uh it's an it's an inherently igopolistic market structure where there are few players that have invested as much as we've invested over as many years as we've invested.
2:52:12
And so we have like 20,000 experts and that number is scaling very quickly.
2:52:17
We can hire like a thousand experts a week.
2:52:18
So when you're asking a model a question about uh the chemical properties of silver or the the history of Sweden or whatever you're asking a model, we have PhDs, masters, experts in every subject that are training these models and there's you know evals and um and and lots of you know inputs into making these models great. >> Got it.
2:52:39
So le leveraging the experience on the training side in order to help the >> deploy. Got it. Got it.
2:52:43
>> deploy. Got it. Got it. And then what and then and then the uh >> the business model for for uh the deployment side is uh just going to repeat it back to you so I so I understand is there's some period that
2:52:59
looks more like services but it becomes a piece of durable software that they can use for a long >> model >> just like talent exactly yes got it cool >> AI AI services is uh exactly that way >> yeah and that puts us in an investing mode we literally invest in our customers. So we put you know field
2:53:16
So we put you know field engineers, field CTOs in with our clients and there's an investment period to build the solution but once the solution's in place uh that is very durable revenue >> and so uh you're you're like a nightmare for someone like an Accenture some of these big consulting companies that don't really have the true expertise around training these models and you get to and and have a lot of organizational bloat. Is that is that right? >> Yes.
2:53:42
So, uh, Matt who, uh, Matt who's our new CEO, we bonded over ancient Greek and Roman history and he calls it the battle of thermopoly because even if you have even if Accenture has like a million people worldwide, has some huge headcount.
2:53:57
If there's three if the pass is only yay wide and there's 300 people at the pass and we're always winning at the pass, then we're going to win every battle.
2:54:04
In other words, if you know, I believe our technology is superior, dramatically superior already to traditional systems integrators.
2:54:10
So, if you're a buyer and you're hearing Accentur's pitch and invisible pitch, which one are you going to pick?
2:54:16
The battle of Thermopathy. Spartans win. >> It's fantastic. >> Spartans are coming. Coming for it. >> What What a market.
2:54:23
This is just a fantastic market. I love congratulations.
2:54:27
Congrats on all the progress.
2:54:29
>> Great to be on the show. Big fan. >> Appreciate it.
2:54:31
>> What's What's a You said prepare prepare to prepare to go public.
2:54:34
What What's the What timeline are we thinking about?
2:54:36
Is that Is that multiple years out?
2:54:39
Is it is it >> we we think in terms of of um you know uh how many hundreds of millions in revenue do you need to have before the story is just obvious to a public markets investor.
2:54:50
Um so we really want to build a you know business for all all markets.
2:54:55
Um so the intrinsic value of the business the ability to survive you know in any condition is key and then we'll take it public. >> Awesome.
2:55:02
All right we'll come back on anytime. This is fun.
2:55:05
>> We'd love to ring the bell to you. >> Great time. Thanks.
2:55:08
Up next, we have Jacob from or uh sorry, David from Juicebox coming into the studio.
2:55:13
I believe in the reream waiting room. >> Bring him in.
2:55:18
>> Juicebox has some fundraising news.
2:55:18
The gong's already warmed up. How you doing? >> Welcome. >> Thank you. Thanks for having me on.
2:55:28
>> David, >> good day today. >> Introduce yourself. Tell us.
2:55:31
>> It picks kind of a good day to drop a cinematic launch video.
2:55:33
>> This is a good day for it.
2:55:33
Uh but take us through it.
2:55:35
uh introduce yourself, the company, the news. >> Yeah.
2:55:39
Uh I'm David, co-founder of Juicebox.
2:55:40
Juicebox is an AI recruiting platform.
2:55:42
Uh we help our customers win the talent war.
2:55:44
Um and we do that by helping them find and engage the best talent.
2:55:47
Uh and yeah, today we're announcing our uh $30 million series A and previously unannounced $6 million seat. >> Boom.
2:55:56
Hit it a few few more times. Few more times. >> There we go. There we go. Congrats. >> Um big day. Okay.
2:56:01
AI uh winning recruiting wars.
2:56:06
What does that what does that mean?
2:56:07
And >> how many how long until you're out of business because no one has a job anymore.
2:56:11
I thought AI was going to get rid of it all the jobs.
2:56:14
>> No silly no silly question.
2:56:14
Uh D are are you working more with uh early stage, growth stage, uh you know, p public companies, all the above.
2:56:24
Like where where are you kind of focused right now?
2:56:27
Because I saw I I I admit I only saw half the launch video.
2:56:30
So you had me up until whatever 50%. >> Nice.
2:56:35
Did you see the the part with the Zach impression?
2:56:38
>> I did see the Zach impression.
2:56:38
Yeah, >> you made it through the important part.
2:56:43
Um yeah, we we work with all kinds of cuh companies ranging from, you know, founders doing their first hire all the way to um kind of growth stage companies, a few Fortune 500 customers as well.
2:56:53
Um some of those companies are Perplexity, um RAMP, uh Kora and and a bunch more.
2:56:59
Uh I think the common thread for where we're able to add the most value is if the right person for the role is actually really hard to find and that can be because they have a specific skill set, a specific set of experiences.
2:57:09
Um something that makes them unique and uniquely positioned to to really excel in that role.
2:57:15
>> I think of hiring as kind of like some sort of chain of events with different point solutions.
2:57:20
There's uh actually finding talent, job platforms, LinkedIn, Indeed.
2:57:25
Then there's, you know, uh the the, you know, you might want to send someone a test or have them fill out a form or process resumes.
2:57:31
You need an applicant tracking system.
2:57:33
I've seen other uh point solutions for, uh, send a video and then ask some questions and then process the video and let the let the user, the hiring manager kind of review videos.
2:57:45
There's a whole bunch of different things.
2:57:47
Do you see yourself as kind of a a a cross crossjourney platform?
2:57:52
Are you focused on a particular landing zone?
2:57:53
like where's the product position today?
2:57:58
>> Yeah, it's a great question.
2:57:58
I think of the the recruiting space as really having um two different ways in which you could add value.
2:58:02
One is um finding net new candidates, people that you wouldn't have discovered or wouldn't have gotten into your pipeline in the first place.
2:58:09
Uh and then everything that's carrying through the pipeline and optimizing that pipeline.
2:58:12
And the latter is usually about like saving time and optimizing conversion rates.
2:58:16
And the former is about um making sure the right talent is speaking to you in the first place.
2:58:22
Um we're focused on that part.
2:58:24
And so our goal is to help companies discover talent uh that they wouldn't otherwise.
2:58:28
And we do that by combining a bunch of different data sources and then using large language models um to run the search and surface those profiles for you.
2:58:35
And so if you think about what like recruiter does today on say LinkedIn recruiter or like a typical search solution, they set a bunch of filters and then start reviewing profiles one by one. Uh it's super manual.
2:58:46
Um requires um a lot of thought still to like think of what could make that profile a fit.
2:58:51
Um, and that's exactly the process that we're able to do with with LLMs.
2:58:56
>> So, is this a direct replacement for LinkedIn recruiter?
2:58:58
Are you are you trying to trying to get people to just be able to uh >> or do you want to like puppeteer LinkedIn with a screen recorder or API or something like that and something like that? I don't know. It depends.
2:59:10
I mean, there's been some companies that have done successfully.
2:59:13
Some of them got acquired by LinkedIn, but I don't know. >> Yeah.
2:59:17
No, no puppeteering of of LinkedIn.
2:59:19
Um, most of our customers are existing LinkedIn customers. Sure.
2:59:22
>> Um our goal is to to help you find net new talent, do so in a more efficient way.
2:59:26
Um so it's really about kind of enhancing the workflow of the recruiter and um enabling to them to do that search that would just be too manual or too timeconuming to do otherwise like looking through 50,000 software engineering profiles is just not something that's like feasible.
2:59:41
Uh but for an LLM it's very feasible and can do it in a couple minutes.
2:59:45
So, uh, there are I mean there are different pools of talent for I mean we do media video production here.
2:59:52
We might want to hire somebody who doesn't really have a LinkedIn presence, but they might have an Instagram account or they might have a personal website where they've done some uh they put up their show reels or what they've worked on.
3:00:01
Uh, do you have crawlers that could go out and find those profiles on the internet?
3:00:06
Like h how how could I think about the surface area of you going out and finding me candidates and sourcing >> that? That's right.
3:00:14
So we look at a bunch of different data sources for different types of roles.
3:00:17
So um I'll give two examples.
3:00:19
One on the engineering side you know GitHub profiles a ton of value on those especially where you know traditional LinkedIn profiles or resumes might be a bit more sparse um especially for for engineers who don't put a ton of info on there.
3:00:32
Um GitHubs can be really rich and a really interesting data source.
3:00:34
Um meanwhile on the sales side often what we see is um you know customers want to find sellers who have experience selling to specific type of buyer persona.
3:00:41
Uh and so there it's actually a bit more about the company data.
3:00:45
Um you know what type of companies does their current company sell into?
3:00:49
Uh what type of a platform are they selling? Is it a SAS product?
3:00:52
Is it a usage based product?
3:00:54
And more and so all of that information we enrich and we try to aggregate upfront.
3:00:57
Uh so we can make those decisions for you um and and help surface the right profiles there.
3:01:02
There's another they put an example they have a tag on on their site called likely to switch which would be like you have a round of layoffs >> funding or vesting timelines.
3:01:12
You know people don't really think about that.
3:01:15
You can see like okay this person's been here >> for four years.
3:01:18
They're probably getting a refresher but they're vested out.
3:01:23
>> Uh that's pretty cool. What do you think?
3:01:26
uh uh do you have a a strong thesis yourself around uh employment uh over the next you know 10 years?
3:01:34
I'm assuming you know Sequoia leading this round.
3:01:40
I I I think that says >> Sequoia confirmed not ag confirmed.
3:01:45
>> Uh no but it says that we're still going to have talent wars in a decade, right?
3:01:50
That's that's what that tells me.
3:01:50
Um but uh how do you think about it?
3:01:55
Yeah, I I think if you're AGI pill, it actually means hiring the right talent is even more important.
3:01:59
Um, and the reason for that is that if you think of like the output of a given person, um, is usually constrained by like their time, um, their previous knowledge and context and then whatever tooling they're using.
3:02:11
And so if the tooling that they're using is getting like exponentially better, um, they use LLM to automate a bunch of the things they'd otherwise do.
3:02:18
They're much more productive than they would usually do.
3:02:21
Um, and that's all a multiple of the person or like the initial person in their kind of raw capacity, having the right person in that role becomes even more important because if AI basically 10xes whatever you do, um, you want to have the highest baseline that you can add that 10x to.
3:02:34
And so it gets even more important to have the right talent on your team.
3:02:38
Uh, and if you think of that in like the context of a software engineer, um, if a 10x software engineer becomes 100x, um, you want to have as many 10x software engineers as you can.
3:02:46
Uh and so my view is that the the race for talent will become more and more competitive as AI gets better and better.
3:02:53
Um because having the right people on your team uh has that outsized impact and I think we're already seeing some of that in like you know the the war for talent and the AI labs.
3:03:01
Um but I think more and more of that will trickle down into the rest of the economy as well. >> Um >> sure.
3:03:09
What do you think about outcomebased pricing?
3:03:10
Famously used in the recruiting industry, most human recruiters, even if they work for a firm, they might get uh one month salary as a bonus if they place someone successfully versus LinkedIn, which is a subscription software completely not outcomebased.
3:03:28
What model do you think will dominate in the future and what are you doing?
3:03:33
>> Yeah, it's a really good question.
3:03:33
Um we so we have like a per seed pricing normal SAS model but then we also have an agent product which is priced per role that the agent works on and so it's not quite outcome based but it's more like usage based how many roles are you deploying the agent for.
3:03:49
Um over time we want to get closer and closer to the outcome based pricing as as is possible.
3:03:54
Um, you know, if you think of recruiting agencies, it's pretty common to do like 20 to 25% fee uh of first year salary, which unlike a dollar value is is much much more than one could probably charge for um a SAS subscription.
3:04:05
And so there's definitely a lot of things that are attractive there.
3:04:09
At at the same time, I think um the market's maybe not quite there yet or not quite ready to uh pay for software on like a say percentage of hire model.
3:04:17
Um though, you know, I hope we'll be able to push in that direction over over time as well. Fantastic.
3:04:24
Well, congratulations on really insane progress. >> Yeah.
3:04:28
>> And congrats on the new round. >> We will talk to you.
3:04:30
>> Thanks for having me on.
3:04:31
>> Have a great rest of your day. >> Thanks for coming on.
3:04:33
>> Uh Sam launched a new feature today uh called Pulse. It's in Chat GBT.
3:04:40
Initially available to pro subscribers.
3:04:42
That's the $200 a month tier, right? Yeah.
3:04:45
>> Um Pulse works for you overnight and keeps thinking about your interests, your connected data, your recent chats, and more.
3:04:51
Every morning you get a custom generate custom generated set of stuff you might be interested in.
3:04:55
Uh it performs super well if you tell chat more about what's important to you.
3:04:59
In regular chat you could mention I like to go visit Bora Bora someday or my kid is six months old and I'm interested in developmental milestones and in the future you might get useful updates.
3:05:10
Think of treating Chachi PT like a super competent personal assistant.
3:05:15
Sometimes you ask for things you need in the moment, but if you share general preferences, it will do a good job for you proactively.
3:05:20
This also points to what I believe is the future of chatbt, a shift from being all reactive to being significantly proactive.
3:05:27
Completely agree with this concept.
3:05:29
Uh, and extremely personalized.
3:05:31
Uh, this is an early look and right now only available to pro subscribers.
3:05:35
We will work hard to improve the quality over time and find a way to bring it to plus subscribers, too.
3:05:39
That sounds compute intensive.
3:05:39
That sounds like they're going to be running essentially a an 03 prolevel query, a chatgpt 5 pro level query every night asking based on what's going on in the news, based on what's what this person searched for, based on all the chats recently, let's summarize and put together some sort of dossier for them and service that.
3:06:00
So that's basically one big, you know, fire up the GPUs every single night.
3:06:05
You got to get the customers to pay for it, at least in the short term.
3:06:09
But what do you think Tyler?
3:06:11
You think you'll use this?
3:06:13
>> Uh, it's pretty interesting.
3:06:13
I I I'm wondering is this the thing that Sam was talking about of the computensive thing they're rolling out?
3:06:19
>> It's he said he said this week we will be launching a couple comput inensive things.
3:06:23
Uh, some of those will probably be announced at dev day.
3:06:25
This is one that he's just announcing this week.
3:06:28
>> This feels like it's it's it's effectively doing something like you know just kind of >> hanging out in the background.
3:06:34
>> hanging out in the background. hanging out in the background p you know somewhat actively somewhat passively just gauging you know what your interests are I can see this as like almost like a having a Google um alert setup >> Jod from Replet says reminds me of Google now but how do I get started I
3:06:51
don't see it in the app so it's obviously rolling uh rolling out slowly >> you can definitely see like an agent uh being built into this like that that seems very natural >> totally >> um >> and and I feel like Chachi PT was already starting to sort of surface this concept of like, hey, do you want me to do this every week for you? Do you want
3:07:07
Do you want me to do this every month for you?
3:07:09
But I haven't actually set up any of those cron jobs and had it stick around and and been become part of my workflow.
3:07:15
But I like the idea of it.
3:07:17
I like the idea of it. And I think that uh this that the big thing is just like for retention if every time you open the chatpt app in the morning instead of just a empty box you're greeted with like here's some interesting stuff >> or you're getting a push notification
3:07:31
>> or a push notification that's going to be huge for retention >> or it's like hey this product launch you want to buy this >> you want to buy this >> uh let's pull up this video from Alexander Wang to share vibes a new feed in the meta AI app >> uh let's see if they beat the uh the slop allegations. I think it's powered by midjourney so it
3:07:51
I think it's powered by midjourney so it looks pretty cool. >> Yeah.
3:07:54
I mean midjourney fantastic model looks pretty good. Yeah.
3:08:00
I wonder how this is actually going to so separate app is what he's saying.
3:08:05
>> It's a new feed in the Meta AI app. >> Okay. In the Meta AI app.
3:08:07
So not not you would expect this to be baked directly into Instagram.
3:08:10
That's probably too much compute out of the gate to just drop it in as a filter or something or button in Instagram cuz if you have a billion people pushing that button, you're just going to get uh get swamped.
3:08:22
I think this beats the uh the slop allegations.
3:08:28
There's some cool stuff you could do with this.
3:08:31
I mean, video generation looks good.
3:08:34
Looks looks V3 level to me. I don't know.
3:08:37
Obviously, these were curated and selected, but >> we'll have to get in the app.
3:08:42
Uh Tyler, before we do the show tomorrow, please use the app for like, I don't know, like 12 or 14 hours, something like that. Just straight. No break. >> Wireheaded. >> Just just lock in. >> Just lock in.
3:08:54
Um >> Max Max uh Conrad in the chat says, "I'm tired of hearing about fundraisers without business metrics.
3:09:02
Topline, bottom line, value created for customers." Max, >> good point.
3:09:07
We're we want to push people harder to say if you want to come on you got to give us some juice >> and if you don't got numbers to share other than the headline >> we should have we should have a very small >> credit to Juice.
3:09:18
I think they've been actually crushing it >> raising money. Yeah.
3:09:21
I mean that is a market where money flows very freely because you're as a business you're ready to sign a check for $200,000 to hire someone is the software engineer.
3:09:32
>> You're like yeah I'll pay $10,000 to get the job done $20,000.
3:09:34
So, the money flows quite freely and that's why there's so many uh executive recruiters, software engineering recruiters that make a ton of money just working basically for themselves just as talent agents moving things around.
3:09:48
>> Um well, Meek Mill and Elizabeth Holmes were going back and forth on X.
3:09:53
Elizabeth says, "Thanks, Meek.
3:09:53
Currently serving a 137month sentence in federal prison.
3:09:57
Would love to work with you on reform.
3:09:58
Here's some legislation I have drafted. Your story inspired me."
3:10:00
And Meek Mill says, "Let me check your story out."
3:10:04
And uh Will Brown uh says, "Elizabeth Holmes, this year is our year."
3:10:09
Elizabeth Holmes, we can never forget this.
3:10:10
Tiger Meek Mill, pray for times like this to rhyme like this. I know. Focus on it, too. >> It's a great one. >> Yeah.
3:10:18
I think they're thinking about different uh aspects of reform. >> Yeah.
3:10:22
>> Um >> well, we don't we don't >> Maybe Meek Mill would want YC arrest where she has to lock in and just repeatedly build the the device forever. >> I like that.
3:10:30
Uh we don't rap, but we do sing. Find your happy place. Find your happy place. Tyler missed it.
3:10:39
>> Give me a second to kind of like out there.
3:10:42
>> Find your happy place. Find your happy place.
3:10:46
>> Book of wanders, inspiring views, hotel great amenities, dreamy beds, top tier cleaning, and 24/7 concier service.
3:10:48
It's >> Well, we got breaking news.
3:10:51
Trump signs order to approve the Tik Tok deal. Avoid the US ban.
3:10:56
The chat was on it way before we were >> um and let's see.
3:11:00
Tik Tok is poised to be spun off into a separate US entity to comply with a 2024 law requiring the uh China par based parent company Bite Dance to divest or face a US ban. >> Mhm.
3:11:14
>> Um uh Trump said, "I had a very good talk with President Xi.
3:11:17
A lot of respect for him.
3:11:19
Hopefully he has a lot of respect for me too."
3:11:21
And we talked about Tik Tok and other things.
3:11:22
We tal but we talked about Tik Tok and he gave us the go-ahad.
3:11:27
Uh, under the deal, a group of US investors, including Oracle and Silverlake, are set to take a majority stake in the new Tik Tok entity.
3:11:33
Let's give it up for Silverlake.
3:11:35
Good to see them throwing some size around. Uh, as well as Oracle.
3:11:40
Bite Dance will maintain less than 20% in equity to comply with the vest or ban law.
3:11:44
Vice President Vance said the company will be valued around 14 billion dollars.
3:11:50
Feels >> low, >> weirdly low. >> Low.
3:11:56
uh >> just the US business. I don't know.
3:11:58
I felt like I mean we talked to Sean Frank and it sounds like they were losing a lot of money on Tik Tok shop and some of the I mean the growth had probably plateaued with >> they were just losing money in general. >> Yeah.
3:12:08
But you got to imagine that that monetizes pretty well. It's such a big network.
3:12:13
So much time on site but yeah it's it's being it's trading like a Snapchat or a Pinterest basically like a non-meta property a non Google YouTube property. Um, >> oh well. >> Yeah, it's crazy.
3:12:27
>> Well, keep checking in on it.
3:12:29
>> So, uh, Tik Tok valued less than Perplexity. >> Okay.
3:12:34
>> Uh, honestly, now I can see why Arvin was trying to get a bid in, you know. >> True. Yeah. So much bigger.
3:12:41
But yeah, basically >> Brett Adcock should have should have taken a crack.
3:12:47
taken a crack. He could have he could have absorbed easily absorbed his market and could have >> true >> a lot of companies that could have absorbed uh >> I'd like to see I'd like to see Palunteer pick them up run the datab 14 14 billion's crazy
3:13:03
>> that is crazy loud >> crazy low uh >> that's I I I'm surprised uh the CCP said yes to this >> that feels >> yeah what's the what's the catch >> I mean >> this feels like it's worth like there's other chips on the table that whe different countries and stakeholders. The UN thing just happened and so
3:13:30
The UN thing just happened and so there's a whole bunch of different pieces and uh the art of the deal is not a single iterated, you know, price a single asset in the moment type thing.
3:13:42
Um could Truth Social have picked it up?
3:13:44
Truth Social's right around there, too.
3:13:46
No, they've uh >> they've fallen.
3:13:47
They're they're in the single digit. Still a unicorn, right?
3:13:50
Our president is still a unicorn tech founder, I believe. >> 4. 7 billion. >> There you go.
3:13:55
He's a unicorn founder, folks.
3:13:57
>> When's uh earnings on November 7th, so tune in.
3:14:00
We got to we got to lock it.
3:14:00
I don't think uh I don't think DJT gives a speech. >> He should. That would be crazy.
3:14:07
>> Uh what you got, Tyler? Uh, do we know?
3:14:09
Um, I I think there are rumors that like the the Tik Tok algorithm would still like be governed by the Chinese, but then it would run on like American hardware or something. Do we know?
3:14:18
>> We're going to do the inference.
3:14:18
We're going to do the inference. They'll do the training.
3:14:20
We'll do the >> inference.
3:14:21
It's like the algorithm with Chinese characteristics. >> Yes.
3:14:24
But uh I am somewhat bullish on the idea of like doing some post training on top of their algorithm or doing some secondary uh like post inference.
3:14:31
you see the ranking of the feed, you're doing analytics on that and seeing like, okay, like there's still some like propaganda stuff here.
3:14:41
We're not >> Is that confirmed that that is what's gonna happen though? >> Uh I don't know.
3:14:45
I don't know if that was confirmed, but uh there are it's not a total black pill if the training still happens there because you can do so much post training.
3:14:53
post training. I I think and uh and monitoring and even just having the data you can start running analytics and understanding like what type of content is showing in what context like are there actual like you know are certain trends or keywords or is there certain
3:15:08
censorship of certain ideas or promotion of certain ideas at least if you have access to the data of what's being served what Tik Tok users are watching you can then understand and then you can apply pressure or you know post training or anything else. Um anyway, in other
3:15:20
Um anyway, in other news, the Lucy nicotine vending machine has arrived at Palunteer. This was a journey.
3:15:28
I worked with Eleano on this.
3:15:31
And John says, "Palanteer has a climate controlled Lucy vending machine for their employees in their office and you're bearish." Uh it was a lot of fun.
3:15:37
So, we'll be sending out a few more of those.
3:15:39
Hopefully getting one in the TBN Ultradome soon.
3:15:40
Um, the last thing was uh the the Financial Times has an article about uh America's biggest corporations keep talking about AI but struggle to explain the upsides.
3:15:50
Um, we'll dig into this another day, but uh it's it's uh they analyzed hundreds of filings that suggest S&P 500 constituents are clearer about the technologies risks than its benefits.
3:16:04
And another one of these data points of Fortune 500 companies being reticent to adopt AI.
3:16:08
I'm not sure if it's just blanket bearish for Fortune 500 companies.
3:16:13
If you're a big enterprise and you're not able to figure out a way to deliver, you know, positive ROI when a magical new technology comes along, uh, it might not be a good sign for you.
3:16:25
Jordy, what else do you want to talk about today?
3:16:26
>> Yeah, I was just confused.
3:16:26
I was trying to dig in and see.
3:16:28
I didn't see A16Z in the announcement anywhere. >> Oh, yeah.
3:16:32
They were supposed to be, right?
3:16:33
Maybe they pulled out or something. I don't know.
3:16:34
I like the idea of the Ben and Mark show just being hardcoded into the post algorithm.
3:16:38
Uh they got Rick Rubin on there.
3:16:41
Show me some Rick Rubin Tik Toks. >> Some Torrenberg.
3:16:45
Little Chris Dixon in there. Some KB.
3:16:47
This would be a good Tik Tok algorithm.
3:16:51
>> They still could be in it, but we'll see what we can ask. We It's It's out.
3:16:55
>> I hope the thumb is on the scale regardless of their ownership position.
3:16:58
>> Putting thumbs on scales.
3:17:00
>> A big fan of thumbs on scales.
3:17:00
Anyways, without further ado, we uh I think got to get on with uh Taipei. >> I think we do. >> So, >> I think we do.
3:17:08
>> We will see you folks tomorrow. Thank you for tuning in.
3:17:09
Leave us a review on your favorite podcast if you're listening.