0:11
[Music] Hey, [Music] hey, [Music] hey. [Music] Come on. Come on.
Search this video
0:30
[Music] [Music] [Music] Yeah. Heat. [Music] Heat. Hey, Heat.
1:15
[Music] [Music] [Music] [Music] [Music] [Music] Hey, hey, hey.
2:39
[Music] [Music] [Music] We came to this world to reach the stars.
3:26
We came to this world to shape our future.
3:37
We came to this world [Music] to reach the stars to shape our future.
3:51
[Music] We came to feel the new music. [Music] Feel the music. [Music] [Music] 3. >> You're watching TVPN.
5:02
Today is Monday, July 28th, 2025.
5:04
We are live from the TBPN Ultra Dome, the temple of technology, the fortress of finance, the capital of capital. That's extremely stupid. You should not do that. It's going to shake up. Which one is it also?
5:16
Now we have a Russian going on.
5:16
Russian roulette going on.
5:22
>> Good morning everyone.
5:23
>> Uh we have a good show for you today.
5:23
A fantastic show for you today.
5:25
We'll kick it off with ramp. com. Time is money. Save both.
5:28
Easy to use corporate cards, bill payments, accounting, and a whole lot more all in one place. Go to ramp. com to get started.
5:34
>> Um speaking of RAMP, >> uh Astronomer is back.
5:37
Astronomer had RAMP on the website.
5:40
Apparently, RAM's a happy customer. Never lost faith. Never lost faith. Yeah.
5:43
Uh so, Astronomer, if you weren't following, if you were living under a rock, uh Apache Airflow as a service, managed uh enterprise SAS platform on top of Apache Airflow for data analytics, streaming data, that type of stuff.
5:57
And they had a absolutely chaotic week last week with their CEO being caught at a Coldplay concert.
6:02
Was that last week or the week before?
6:05
>> I think it was might have been the week before.
6:07
>> The week before, but had to have been.
6:08
Um, the CEO was caught having an affair at a Coldplay concert.
6:11
Chris Martin called him out on stage and said, "Oh, those people look like they're having an affair."
6:17
>> Was Was Did he actually say that live?
6:19
>> He did say that on the video.
6:19
So, they're panning around to the different kiss cams or different just cameras and they spot uh the CEO of astronomer hugging the head of HR. Yep.
6:28
Can't even hug your employees anymore apparently.
6:33
>> Can't even give your chief people officer a hug in this country anymore.
6:36
And once they see themselves on the screen at the show, they recoil in horror >> and turn around and uh and the other person in HR who's sitting next to them is like, "Oh my god, what's going on?"
6:47
>> Well, that's what I didn't understand.
6:49
Was that actually >> apparently people people went and dug it up and found out that she works there, too.
6:54
>> I thought that was people just saying, "Hey, this person looks like this person."
6:59
>> It seems it seems wild that the HR department was hitting the concert with the CEO.
7:04
>> It was like videos open secret.
7:04
I don't Anyway, it did not it was not good for astronomer.
7:09
Uh we were discussing this.
7:11
Lulu was saying the CEO's got to go because he's a hired gun, not a founder and uh it's it just displays very bad character and it reflects poorly on the company.
7:19
We were kind of going back and forth on this as as the like about the idea of like okay yeah like the CEO did something bad in his personal life but like do you really want to find an alternative to your managed Apache Airflow service?
7:32
like it's kind of a hassle to rip that out if you're happy with the product.
7:37
>> To be clear, we we said that the company would be fine.
7:39
They did need a new CEO immediately and the founder, I believe his name is Pete, uh stepped up within days. Pete dejoy. >> Oh, wait.
7:49
So, founder's back in the CEO.
7:51
Back in the >> That's the That's the real bull case for the Ashton Hall sound effect.
7:55
I want to hear some good news.
7:59
>> Astronomer is now founder mode, everybody.
8:00
Astronomers is totally in founder mode.
8:03
>> They were already delivering the world's data. >> Yes.
8:06
>> And now they're in founder mode.
8:07
>> They're in founder mode.
8:07
I think they're uncertain. I'm excited.
8:08
Another billion dollars for Bane Capital.
8:11
Let's hear it for Bang Capital.
8:12
They haven't got enough index. >> An index. Nice. I didn't know that.
8:17
>> Anyway, so >> and I actually I I emailed briefly uh with Pete to Joy and he said he's a fan of the show. Amazing.
8:23
We're hoping to get him on.
8:26
We don't want to actually talk about any of this stuff.
8:28
want to talk about Apache Airflow.
8:29
I'm so into Apache Airflow now. I'm so ready.
8:35
>> Um, but anyways, so late Friday night, uh, they uh, astronomer >> clearly they didn't want they didn't want us to react to this on stream, so they put it up after we after we logged on.
8:44
>> Let's pull up the video.
8:45
>> I want to watch the full thing.
8:47
>> Thank you for your interest in Astronomer. Hi, I'm Gwyneth Paltro.
8:48
I've been hired on a very temporary basis to speak on behalf of the 300 plus employees at Astronomer.
8:56
Astronomer has gotten a lot of questions over the last few days >> and they wanted me to answer the most common ones.
9:07
>> Yes, Astronomer is the best place to run Apache Airflow, unifying the experience of running data, ML, and AI pipelines at scale.
9:15
We've been thrilled so many people have a newfound interest in data workflow automation.
9:20
As for the other questions we've received, yes, there is still room available at our Beyond Analytics event in September.
9:31
We will now be returning to you know game changing results is Chris Martin's ex-wife. >> Yes.
9:37
>> Thank you for your entire thing. >> It's very funny.
9:40
It's it's not really like Chris.
9:43
It's not like getting back at Chris Martin in any sort of weird way.
9:46
It's just funny that it's like another voice from that universe really. >> Yeah.
9:50
>> Yeah. that the so they they remain close friends and co-parents in in in so many ways like she's like the best spokesperson for this because >> anyways you know >> remarkable how quickly they shot that you know they had to like probably like
10:06
a conference room near her office or like her house or something like that >> perfect response you don't need from astronomer there at all >> six lines it's really >> ties into the story it made sense it wasn't just some random celebrity that had some funny tiein. >> Yep. Yep. Yep. >> Yep. Yep. Yep.
10:21
>> And I think it was incredibly well done. Uh Lulu broke it down.
10:23
She said, uh, "Laughing at themselves was the right move because humor does four crucial things.
10:29
Connects with a new audience.
10:32
>> Even for people who don't care about Apache Airflow, being in on the joke together forms a connection with astronomer.
10:37
Diffuse tension by joining the ridicle ridicule.
10:39
They're no longer its subject or its object. Get closure. They said it out loud. The joke is tapped. Everyone can move on." Yep. >> Signal a fresh start.
10:47
the CEO and HR lady are gone.
10:49
It's a new management team and making light of this shows they're they've consciously uncoupled from the past.
10:56
So great love it great breakdown.
10:59
>> Very well written, Lulu.
10:59
Uh the the third point is so funny because there's a little bit about like like nothing will kill a joke like independent of all the crazy astronomer Chris Martin Coldplay thing.
11:10
There's nothing that will kill a joke faster than a series D enterprise SAS company making the joke.
11:16
And so if they're making if they're jumping in on the joke, it's like, well, we wanted to kill the joke.
11:19
We wanted the joke to stop. And so we jumped in. We're playing along.
11:23
And we got the last literally the last laugh.
11:25
Like like if anyone tried to post something about astronomer CEO Coldplay, all like they would immediately be everyone would be like, "Yeah, we've moved on."
11:34
>> The real question is, is the IPO windows open?
11:37
Should they go public right now?
11:40
>> Memeto >> and actually have Gwennneth step in like kind of like chairman type role, you know, really expand the role, not just kind of price. >> They need a treasury.
11:49
>> They do, >> but they need something that that that speaks to the the the history of the company.
11:54
They need to buy some some funny funny asset to put on the balance sheet.
11:58
I was thinking about um they we've moved on from Bitcoin treasuries to like the further out risk curve ones.
12:07
>> Well, there's just all the Ethereum treasuries, >> GameStop treasury, which is hilarious.
12:09
I I think the next generation is just straight up lottery ticket treasury. Just just scratchers.
12:18
>> I thought you were going to say like they should put some like match group on the balance sheet.
12:22
>> They should that would be good.
12:22
That would be more like tied to this.
12:23
And I think that makes sense for astronomer.
12:26
Get a bunch of match group.
12:29
get a bunch of match group stock on your balance sheet.
12:30
>> By the way, we're >> or like uh uh Eventbrite maybe or who who who runs like the Coldplay concerts?
12:38
It's like it's like uh uh didn't Taylor Swift like sue them? Ticket Master.
12:40
Get some Ticket Master stock on the balance sheet, you know? Who knows?
12:44
That would tie a little like the values of the company. Exactly. Some potential. >> What was that?
12:51
There was that biootherrapies company that was saying like we're fighting financial fraud by buying GameStop stuff or something like that or financial inefficiencies.
12:58
But yeah, get some scratch market.
13:02
>> Get some scratchers on the balance sheet for sure.
13:04
Get some lottery tickets.
13:06
Everyone's like, "Yeah, they have they have $50,000 in lottery tickets, but if it hits, this could be $500 million on the balance sheet."
13:14
Yeah, it's it is such a I think a year from now we'll look back and say like they found a way to actually turn this into a win. >> Totally.
13:25
>> Now some people >> if if you know this is maybe some of the best crisis comm's work we've seen y >> in uh you know this decade but >> I think if you you now have a million millions of people that like have think your company is kind of like cool and funny >> and they're aware of it.
13:43
they're aware of it >> and it would have cost them it would have cost them I'm sure this Gwennneth video cost millions of dollars.
13:48
I think it probably would have cost them to try to to try to build that type of brand recognition otherwise >> like just just traditionally would have cost multiples of that.
14:02
>> I only have one note on the video.
14:02
I mean Autism Capital here says you have to give credit where credit is due.
14:06
This is 10 out of 10 PR recovery.
14:08
I think it loses one point because it was hard posted and not rereamed.
14:12
one live stream, 30 destinations, multiream, and reach your audience wherever they are.
14:17
They should have rereamed it anyway.
14:17
Um, >> fantastic comeback.
14:20
Uh, Lulu says, "Next move is for an astronomer competitor to hire Chris Martin to do a video on how their product is the best at helping you gain visibility in any environment and keep your private networking secure."
14:36
Uh, and and uh, somebody in the comments says, "The chain is going to end with Brad." >> This is the thing.
14:40
I don't even know who astronomers competitors are.
14:41
don't know who their competitors are, but I don't know, maybe they should.
14:46
>> That's why this is a win for Astronomer.
14:48
>> It's a huge win for Astronomer.
14:48
Um, and yeah, I mean, also interesting because this kind of plays into what we were talking about with with Paul from browserbased, this idea of like the like Apache Airflow is open source.
14:58
You would expect that managed Airflow would be something that's, you know, totally in AWS's wheelhouse.
15:04
And yet they were able to scale to a series D company, 300 employees, like clearly doing well.
15:09
Um and now they have this like breakout moment and u they still are probably facing fierce competition from the from the hyperscalers but um and from the from the big uh cloud platforms but they just don't >> it's so the timing of this is so insane. Bane led the series D.
15:25
It was announced on May 1st. >> Let's go.
15:29
>> So >> love it >> just very recently here.
15:31
I mean they've been putting up some incredible numbers.
15:35
>> Let's get Mitt Romney on the show.
15:35
have him talk about >> last f the last fiscal year astronomers saw 150% year-over-year ARR growth worldclass 130% net revenue retention and 90% product utilization with customers so let's see >> I mean I think they're going to have a massive uh end end to the year >> speaking of high NPS products let's tell you about figma.
15:55
com think bigger build faster figma helps design and development teams build great products together you can get started for free at figma.
16:02
com >> and we will be in the great city of New uh this week. Yeah.
16:07
>> For the Figma IPO, we will be live from NY, the New York Stock Exchange on Thursday.
16:13
>> That's what the cool kids call it, NY. >> Nice.
16:16
>> I always just call it NYSE or like the New York Stock Exchange.
16:18
But when you're saying it every other word because you're in that world, when you're big in that world, you don't >> when you're taking when you're taking companies public like every other week. >> Exactly. Exactly.
16:28
You gota you gota use the cool what the cool kids use. Uh okay.
16:32
bull or bare case for Figma.
16:32
I go to Figma make and I tell it build me a collaborative design tool. >> Don't make mistakes. >> Don't make mistakes.
16:42
Recursive the snake eating its tail.
16:45
You use Figma to make Figma, then you don't need Figma anymore. Is that a bare case? What's going on here? >> Well, I think so.
16:51
You're basically still paying Figma to host. >> Oh, okay. Okay.
16:55
So, they get you to host it. Okay.
16:57
That's how they get you locked in. >> You're paying.
17:00
>> That's how they get you locked in.
17:00
>> You're paying one way or another. >> Yeah. I like that though.
17:02
It's a it's an extens it's an existential risk for all these platforms that allow you to build software vibe code.
17:08
>> That's why every SAS >> first thing I want to vibe code is a vibe coding platform. >> Well, yeah.
17:11
Every that's why every SAS company has to have a vibe coding product now. >> Yes.
17:16
So you can vibe code the product itself. >> Yeah. >> Yes.
17:19
The the the toological vibe code.
17:21
>> No, that is I mean so the question people have been saying like okay at at some point in the future you'll be able to oneshot products. >> Yep.
17:30
And I I would say I have uh really strong conviction that in the next few years you'll be able to oneshot a design tool. >> Yep.
17:39
>> Will you be able to oneshot a design tool for that works in the enterprise that work as like when you think Figma has been like shipping features every single day for a decade now. Yep.
17:49
And um and even if you knew exactly which features mattered and how they all work together um it would be very difficult to uh create a onetoone clone.
18:04
>> And then also there's the network of if you hire a designer and you're like hey >> I need you to use my >> use my vibe coded Figma knockoff that I vibe coded in 20 bucks a month. >> Yeah. >> Yeah. Yeah. Yeah. For sure. That Yeah. And and then Yeah.
18:19
So the ecosystem is very very important and also >> and there's a whole app ecosystem. >> Exactly.
18:24
>> It makes me you know it definitely makes me more bullish on companies that have these like developer app ecosystems. >> Yeah. Yeah.
18:32
>> Um I mean Shopify is the same way.
18:32
You can one maybe you could oneshot like an e-commerce storefront product but can you >> are you then going to oneshot the downstream?
18:41
downstream? I mean I as soon as as soon as LLMs were writing code and we were talking about like AGI takeoff and super intelligence I had this like running thought about um okay so at a certain point you can go to an LLM or a vibe coding platform and say like build me an
18:58
e-commerce website and it will just say like okay setting up Shopify but in the far far future it could just say okay applying for a banking license applying for a money transfer license I'm going to rebuild Stripe I'm going to rebuild Shopify. I'm going to rebuild a database
19:12
I'm going to rebuild a database from first principles.
19:14
I'm going to use just raw >> make a data center.
19:19
>> I'm going to make I'm going to build >> if I want to build an internet company, I should have at least one data center.
19:23
>> And it all just does that in one prompt.
19:25
>> Um because one prompt fires off I mean h how many manh hours have gone into building Stripe or building any of these companies?
19:32
It's like you know tens of thousands of employees for most of them for you know decades.
19:36
You add all that together.
19:38
But if the LLM can do that in the if the AI system can do that in the data center in just a few minutes in hyper compression, who knows?
19:45
Maybe uh Logan Bartlett has another take on the astronomer video.
19:49
He says, "The astronomer video is great on its own, but I'm even more impressed the leadership team and the board were able to come to consensus to make this investment and take this risk absent the CEO they've had for two year for the last two years.
20:02
I can't imagine everyone was on board with this initially.
20:04
So major kudos to everyone getting there eventually and taking this chance.
20:08
This bodess well for their future IMO and I agree uh like even with Gwennneth Paltro it it feels so funny but like there's this idea of like let's let's put out not a standard legal statement is feels risky and it's so easy for someone to step up and say hey like let's not take this risk. It's not worth it.
20:30
Well, they put out the quick statement that Pete was stepping back into the >> Yeah, they had put out a few statements, but clearly um something something was was clicking. >> Question.
20:39
What should Andy Byron, the CEO, uh having the affair?
20:42
Which what >> I mean, that's the best part about this video is that it doesn't take shots at him.
20:46
It doesn't it doesn't it doesn't punch down.
20:47
It doesn't make it like, oh, it doesn't drag that in.
20:49
It doesn't make it more complicated.
20:51
And a lot of times when when CEOs do get pushed out, um there's like lawsuits about comp and was it a fair to release people.
20:59
And so there's like all these things that can come back.
21:03
Like if you are especially if you're a founder and you're going back into a company where you've hired a CEO, uh you should you should probably not be talking about that C that CEO.
21:12
uh because if you come out and say they were not good and that's why I had to step back in then that could hurt their career prospects and then they could sue you for defamation or something like that.
21:24
So there's a lot of risk to anything around that all the corporate coms like it is like the lawyers are like annoying but like they do make a good point that like there is financial impact if you get it wrong. So uh very very good.
21:35
Um, and then, uh, STA stays says, uh, I think we're going to find out that Chris Martin felt bad, asked Gwennneth Paltro to help out, and they gave astronomer an offer to do damage control.
21:47
Uh, my guess is that this is entertainment industry magic happening, not data tech magic.
21:51
H Logan says >> Chris Martin wants to wants CEOs who are having affairs to feel welcome at his romantic concerts.
21:59
He's like, this could be really bad for business.
22:01
This could be bad for ticket sales.
22:03
I have to go into damage control.
22:04
I don't think Chris Martin's behind this.
22:06
This is a ridiculous theory that this stay say stay sassy or whatever.
22:10
Uh Logan says totally possible.
22:13
>> I do I think there could be some I think there could be something here.
22:16
I mean >> I don't think so. I don't know.
22:17
Like what why he feels bad that this just happened at his show.
22:22
>> Chris Martin has the data.
22:22
He might be like 10% of people at my concerts are having affairs.
22:27
>> You think he's storing Apache Airflow >> or Yeah.
22:29
Maybe maybe he just really cares about >> ticket sales are going through Apache Airflow and he's monitoring it using astronomer.
22:34
He's just like, "No, my bags. I love this company."
22:36
He's like actually accidentally an investor as well through some fund.
22:41
Maybe he's Bane Capital LP. Who knows? He might be at index.
22:45
>> He might be heavily anchor GP >> anchor LP in index >> Bane and index.
22:52
>> Um, and to close out, Lulu said, for everyone asking me, this wasn't me.
22:55
I do not work with astronomer, but I think it was very well done. Kudos to their team.
23:01
And it says a lot about Lulu's brand that whenever good PR happens, people are like, "Lu has to be behind this.
23:06
It's it's it can only be her.
23:08
>> It can't scream Lulu." >> It did. It did. It did scream Lulu.
23:12
>> I didn't I didn't um I didn't ask her because I didn't want to know because like I just like, you know, >> like the mystery.
23:18
>> I like the mystery of it.
23:18
I feel like if I knew I' should probably say something, you know. >> Yeah. Yeah.
23:23
Anyway, uh let's shift gears.
23:25
Let's tell you about Vanta.
23:27
Automate compliance, manage risk, improve trust continuously.
23:29
Vantis trust management platform takes the manual work out of your security and compliance process and replaces it with continuous automation.
23:36
Whether you're pursuing your first framework or managing a complex program, uh, >> cheers to Vant.
23:42
>> In the other side of PR statements, the T app hack is an absolute disaster.
23:45
Uh, this is >> and their statement was a disaster.
23:51
>> Their statement was a disaster.
23:51
It does not seem >> uh somebody just informed me a friend of the show that it was Ryan Reynolds agency that pulled it off. >> No way. >> Makes total sense. >> Yeah.
24:01
Ryan Reynolds >> the bridge between >> tech Hollywood tech.
24:05
>> Oh, he's a master master of craft.
24:05
And I feel like he's done a number of those like solo director reported.
24:09
So there's an article Ryan Reynolds maximum effort ad agency turned Astronomyl's viral moment into marketing gold. >> Wow.
24:19
The production company was involved with the latest astronomer video featuring Gwennneth Paltro.
24:23
The ad is being hailed as a master class in Crisis PR. I >> agree.
24:28
Um, >> astronomer Fortune is saying astronomer got the last laugh. >> They did. They really did.
24:36
>> Releasing it late on a Friday, too, is great.
24:38
>> Shut shutting down the work week.
24:40
>> Well, you don't want AWS to trade down too heavily on the news.
24:42
You know, if you release that during market hours, it could be turmoil.
24:46
could trigger an entire market selloff.
24:47
And last week, we had five back-to-back all-time highs in the S&P 500.
24:52
That's why we're wearing white suits.
24:54
>> And we're wearing white suits because we have peace with Europe. >> Peace with Europe.
25:00
>> We'll get to in a little bit.
25:00
But let's talk about tea.
25:02
They released a statement.
25:05
>> The tea app is an app that allows women to report red flags about men they are dating.
25:09
It's based loosely on these are we dating the same man Reddits.
25:11
And it's not been handling its crisis well.
25:14
I don't know if you want to read the official statement. >> Yeah. So lawyer voice 44 a. m. PST on 7:25.
25:22
We identified an unauthorized access to our systems and immediately launched a full investigation with assistance from external cyber security experts to understand the scope and impact of the incident.
25:31
Here's what we know at this time.
25:34
A legacy data storage system was compromised resulting in unauthorized access to a data set from prior to February 2024. Yeah.
25:44
And and so I think people were already kind of like factchecking this. >> Yeah.
25:50
>> Because there's the date they said >> because I remember you told me that >> there was like 60 million million users or something and it was at the top of the app store.
25:57
So you would assume that it it was it was at this crazy viral inflection point and the amount of data that they were onboarding just in the past week was probably immense.
26:04
But prior to February 2024, that's a long time ago.
26:09
That's what 15 16 months ago.
26:12
uh pretty pretty distant.
26:12
So maybe it is possible it was leg legacy data storage, but there's more posts that we're going to go into that kind of illuminate exactly what happened.
26:21
But they say that the data set included 72,000 images, including 13,000 selfies and photo identification, including 13,000 selfies and photo identification submitted. What? What?
26:37
selfies and photo identification.
26:37
Like that's not like photo IDs, photo identification images.
26:41
Like that's not the correct phrasing for that.
26:43
Um submitted by users during all the reporting is that there's 13,000 photos leaked. >> Okay.
26:51
So like 13,000 people got their stuff leaked basically.
26:53
Um >> which feels much lower than than I think what what people had.
26:58
assumed if you had downloaded the app at any point in time, everything was leaked in that in that data data set.
27:06
But anyway, uh let's go into the PR responsib I mean it wasn't like they were hacked.
27:14
They just like >> they made there was something that was publicly accessible.
27:18
>> Yes, they apparently they were using a Google Firebase database which is um one of these kind of easy to spin up backends.
27:25
Tyler, have you used it before? Firebase?
27:27
I I actually used um in high school like the first ever mobile app I made I used Firebase cuz you can basically do >> back when you were a real rookie.
27:37
>> You can do O you can do like normal databases you can do like image storing you have like everything in one platform. >> Got it. Yeah.
27:43
And it's just like but it specifically Firebase is supposed to be designed as a database that you can access from the front end.
27:50
So I believe the like most of the value prop is that you you you can focus a lot more on frontend coding and still access data on the back end and you kind of get all the routes out of the box or something.
28:06
>> Yeah, I think that's probably true.
28:06
It's just like it's like very simple to use. Sure. It's very easy.
28:10
It's very easy to use with mobile. Yeah.
28:12
>> Um >> is it a competitor to Parse?
28:13
I think I think uh Ilia over at Matrix started Parse and sold that to Facebook and then I think Firebase was kind of the the Google answer to to Parse.
28:21
Google answer to to Parse. But the these ideas of like quick, easy to use mobile backends, not nearly as robust or scalable as something that's, you know, like a true enter enterprise like AWS installation, but something that's um
28:36
certainly usable if you're trying to get up quickly, which kind of makes sense that if they launched this back in February of 2024 that they would have this quick and simple database and then once they scaled, they might have moved to something else. So, I don't know that
28:47
So, I don't know that that's wrong.
28:49
It's just kind of the way that they wrote this was not great. So Lulu breaks it down.
28:53
She says, "Eight things to note in the statement that T finally released about their data breach. No apology really.
28:58
They should have apologized for sure.
29:00
Two, it dodges responsibility.
29:03
They say legacy data storage systems, corporate and it's fundamentally dangerous to their users, right?
29:10
If if if if random people online can identify this like single woman who's located at this address totally in this city >> totally >> like that that's like like the fact like the the whole response was terrible.
29:25
Um and the level >> it's so much different from astronomer because like astronomer you see the CEO doing something bad you're like well he didn't leak my data he didn't leak my API key like it would be honestly worse hurt any of their users.
29:40
Users are like, "The product's solid.
29:42
>> The product's solid recommend." Yeah, exactly.
29:45
>> But this is like incredible incompetence to put personally identifiable information in a publicly a publicly accessible database. Yep.
29:53
And then say, "Oh, uh, we were hacked." >> Yep.
29:57
>> And don't even and not even say sorry when you've you've risked the safety of your users. Yep.
30:02
>> And it's embarrassing.
30:02
deeply embarrassing for these users.
30:04
People are turning it into a um >> like King of the Hill, whatever.
30:10
>> It was like Hot or Not >> Hot or Not with like a with like a a leaderboard.
30:14
And then they also built just a Google map that you could look at and see where everyone like the pins on the map.
30:20
Uh just completely extreme.
30:20
Way worse in my opinion than even having my credit card leaked.
30:26
You know, this is like this is a dating site almost or like it's dating site adjacent and so it reveals something. cancel a credit card.
30:33
You can't cancel the fact that your >> ID >> and and everyone everyone has a credit card or it's not it's not some like secret thing, but like this question of like were you checking on someone you were dating?
30:45
It's like were did you have a trust issue with someone in your relationship that reflects something about you and that's like something that people probably wouldn't want to share at all.
30:53
Uh especially not considering the vibe on the 4chan thread that was actually the precursor to all of this. Yeah.
31:01
So, >> it's like, hey, we shared your we shared your name, date of birth, >> address, driver's license number, >> but they didn't get your email. >> Yeah. Yeah. The the government IDs.
31:12
I hadn't seen anyone actually prove that that happened that because it was unclear like the IDs went out there.
31:22
Maybe that's just locked down.
31:23
>> There was pictures of the IDs floating around. >> Okay. Rough, rough, rough.
31:25
Um, and so >> it was so I mean the the timing was was wild too because last week I I posted um I posted founder of T dating was uh previously a product leader at Salesforce.
31:40
Massive moment for big tech PMs. Never doubt them again.
31:42
PMs. Never doubt them again. obviously like poking fun at you know just like the idea that like you know people love to say like oh big tech PMs are just changing like the color of a button you know >> getting to the top of the app store is impressive whether you're Nikita Beer or
31:58
some college kid hacking an app together or a big tech BM or you know Mark Zuckerberg it's always it's always hard it's a knockout dragout fight on those leaderboards >> it's quick to say I spoke too soon >> this uh but uh >> how many likes did the I spoke too soon get >> 10k head. >> Yeah. Always dunk on yourself. Don't let >> Yeah. Always dunk on yourself.
32:15
Don't let other people >> Everybody uh we can go back to doubting big pe big tech P V P V P V P V P V P V P V P V P V P VMs and then until they can get back to the >> Yes.
32:24
Um but you got to review your code.
32:28
You got to do it on graphite. dev.
32:28
Code review for the age of AIO.
32:30
Graphite helps teams on GitHub.
32:32
I don't want to say the graphite would have prevented the situation, but >> but it might have. >> It might have.
32:39
This is not a direct recommendation, but Graphite helps teams on GitHub ship higher quality software faster. You should be on GitHub.
32:45
You should be using Graphite.
32:46
You should be reviewing all the code for security purposes.
32:48
So, >> it would have been cooler if they came out and they were like, "Yeah, we never would review code.
32:53
We actually wouldn't we wouldn't actually we like reviewing code like that's why why look back when you could look just vibe code the future?"
33:04
>> Seems like the future.
33:04
Uh so Lulu says uh it's an obvious lie that they say we have no evidence to suggest that photos can be linked to specific users.
33:12
Photo identification is the definition of photos linked to specific users.
33:16
This sentence sentence almost felt too s too stupid to type too slow.
33:21
The internet has already memed you five feet into the ground yet only now are you issuing a statement and it still looks rushed.
33:28
The information here is still minimal and inconsistent.
33:31
And the statement is rife with run-on sentences and other syntax errors. Brutal.
33:35
Uh Weasley passing of the box, saying the data was stored in accordance with law enforcement requirements related to cyber bullying investigations implies that law enforcement is to blame for their negligence.
33:46
High fluff ratio uh with empty euphemisms like robust and secure solution.
33:51
These words inform us of nothing.
33:53
It's a accountability theater.
33:55
Yeah, they should just say like, hey, we don't use, you know, public Firebase buckets anymore.
33:59
We've moved on to Postgress with a lot of security and two-factor authentication or something.
34:06
>> We use secure databases now.
34:08
>> We used to not use them. >> Be Yeah, be be much.
34:10
Here here's here's the question I have is um >> however many women were impacted um I imagine they will it's pretty easy to identify who they are and I imagine lawyers are reaching out to them and saying let's file a class action against TAP >> uh for damages.
34:27
I think there's real damages here right these it's like public embarrassment safety concerns etc.
34:35
Um, and uh it's maybe hard to prove like direct monetary damages, but it looks bad.
34:42
There's definitely some some uh some type of case here.
34:44
And so uh however well the T app is doing in the app store. >> Oh yeah. Is it still up? We got to look that up.
34:52
Uh you look up the T app rankings.
34:54
I will keep reading from Lulu.
34:57
She says >> still number one in lifestyle. >> Nice.
35:00
stiff legal ease written by family lawyers. Four and a half stars.
35:05
>> She says she's confident that uh blank is the of the utmost importance to us is not a sentence that has ever been spontaneously said out loud by a real person in human history and not taking lessons from previous data breaches.
35:19
There are so many cases case studies of bad statements including crow crowd strikes below. Learn from them.
35:23
So chatbt is back at number one.
35:27
Uh, t- dating is still at number two. >> Wow.
35:32
>> To have that viral of a uh of a hack as they called it or just a a release of your users's data as some might call it.
35:41
>> At the same time, you know, you're going viral for getting hacked.
35:42
You're also, you know, driving a ton of attention and people are like, "Okay, I will >> I better check it out.
35:48
>> Yeah, I better check it out."
35:48
Or or I'll take my selfie with professional lighting.
35:52
So, if it lock if it if it leaks, >> I want to be at the top of the hotter nut.
35:56
>> That was my That was my lesson.
35:56
I was looking at some of the selfies and I was just like like people are dunking on these for a variety of reasons, but the real lesson is like if you ever have to take a verification photo for a selfie, get get a nice soft box going.
36:07
Get put on some wear a suit, put on some makeup, make sure make sure you get some powder. Yeah, exactly.
36:13
You want to be looking good.
36:15
If you're taking the the future leaked selfies so that you look you look fantastic.
36:19
The there was a guy who got a mug shot and became a model.
36:22
You heard about this guy?
36:23
14 for uh 2014, a guy was a I think he was a member of the Crips.
36:29
He uh he was arrested on some I think legitimate charges. He was sent to jail.
36:34
Um but his he was absolutely serving in his in his uh in his mug shot. You know this guy?
36:42
>> I No, I just never would have thought I'd hear >> hot mug shot guy.
36:45
Guarantee you'll find it.
36:47
>> I I' never thought I'd hear you say the word serving.
36:50
No, he really like it was just like looks like uh Tyler Cosgrove in the upcoming ad.
36:55
Looks uh looks like future male model and it was and he became a future >> after this.
37:00
So we shot an ad with uh and Tyler's the star. We're not in it at all.
37:05
And after this drops, if >> look at this guy, look at this guy. >> He's serving. >> He's serving. Right.
37:09
So this went super viral and every and all these all like tons of people were like, "This guy's so hot." Blah blah blah.
37:15
He became an actual model and he and he married or had kids with or dated like the heir to a multi-billion dollar fortune. >> Never give up. >> Never give up. >> Never give up. >> Never give up.
37:28
You're just one mug shot away from fame. >> Yeah. >> Yeah. Better.
37:32
>> Tyler, if physics doesn't work out, maybe crime and then mug shots and then >> don't hurt any Don't do do petty crime. >> Victimless crime.
37:41
>> The best kind of crime. >> Yeah. Victimless crime.
37:42
like go to a CVS and break the glass and get like a one stick of deodorant or stand outside and get arrested >> but look really good >> and then give the deodorant back. >> Yeah.
37:54
Have a little pump going be in shape when you take the mug shot so you look great.
37:58
But uh back to the tea app.
38:00
We were talking about how much money they were making.
38:02
And this is what this is apparently from nosy bystanders. I got the tea app.
38:07
Who y'all looking for in what city? Charging $1 a search. Okay. Okay. Okay.
38:13
I thought I thought that noisy nosy bystanders was saying that that every time you searched on the tea app, it was $1, which would be crazy monetization, but I guess nosy bystanders saying that they will search for you for a dollar.
38:25
Um >> yeah, look up uh your the whatever whatever dudes >> Yeah.
38:31
>> you know, you're interested in learning about >> looking at these dudes.
38:33
I like this guy with the basketball.
38:35
He's looks like a good dude, you know. >> I don't know.
38:38
I'll reserve uh I'll reserve my judgment once I see his full profile.
38:42
What I don't understand is how do they how do they source those images?
38:47
>> I I I think the I think the users upload them.
38:49
So I go there and I say I had a bad experience.
38:53
>> This makes me think this makes me think just burn the whole app down because >> because going and and taking >> Yep. >> like Yeah. I don't know.
39:02
>> So this was the debate that was happening on hacker news that I kind of scrolled through.
39:06
So the the pro tapp camp would say that this is about safety and that this app is effective for um identifying and like abusive men.
39:16
And so a woman uh is in a relationship with a man. The man hits the woman.
39:23
She leaves the relationship and she wants to let other women know that hey this guy might hit you if you get into a relationship with him.
39:31
So she gets on the app and says I was dating Bob. Bob hit me.
39:33
and it's all true and that acts as a warning and so that's increasing the amount of safety in like the dating marketplace.
39:42
Now the flip side is that uh >> somebody could get their feelings hurt and do the same So, so people could lie and say this person hit me when they didn't because there isn't the abundance preponderance of evidence that is required in the courts.
39:55
But then also you on the T app you could put red flags that weren't actually related to things that are uh like like illegal or deemed by society to be that bad.
40:05
So uh the flaw that like the flags could be like ghosted me which is like impolite but not not at the same level as domestic abuse clearly.
40:16
And one of the flags apparently was just is bald.
40:22
>> So it's like if a guy shows up at toupe and you get rugged, no pun intended.
40:27
>> Hey, let's uh >> That's wrong. That's wrong.
40:31
>> Let's support our bald our bald brethren.
40:33
You got Jeff Bezos, Mark Andre.
40:35
You got some of the absolute boys.
40:38
>> And uh Charlie X >> Charlie XCX. >> Yeah.
40:42
She just she just married a bald guy.
40:43
>> She just married a bald guy. >> Married a bald guy. Great. I love the Bald. >> Yeah.
40:48
Some of the greatest men in history were >> And so and so the point is is that like there's this question of like is it an app for increasing safety?
40:54
Is it an app for just talking trash and being mean?
40:58
And that's kind of the debating line.
41:00
And no one really knows what percentage of reports were real and what percentage of reports were actually based on uh like things that truly cross the line.
41:09
Because if if the app is 99% people complaining about people showing up to dates with two pays when they're secretly bald, like that's kind of ridiculous.
41:17
And I guess you are getting catfished in one some way, but it just doesn't feel >> he has a flight to Turkey plan. >> Exactly.
41:23
>> What if it's on the road map? >> Exactly. Exactly.
41:25
So >> could it didn't even give him a chance because >> Yeah.
41:29
And so and so there's a there's a world where some some pieces of the app were very helpful and they increased safety and there's others where they were basically just like cyber bullying each other and it was all just a mess.
41:39
Um either way uh all of that >> that starts with good security on the app.
41:44
Like you can't even have that conversation when you're just leaking everyone's data. >> Yeah.
41:48
The uh there was an app um I forget what it was called but I anytime you have these like anonymous apps >> Yeah.
41:56
They're just prime for cyber bullying.
41:59
>> There's a huge trend of this.
41:59
Uh they used to go viral at South by Southwest post social networking boom.
42:04
So Twitter originally went viral at South by Southwest.
42:09
I think Jack Dorsey, Bis Stone, they stood up on stage and they said, "Hey, we're launching this new app. It's Twitter.
42:14
You can text this number and whatever you tweeted in that conference room would go up on the on the board that they were live streaming basically."
42:22
And it was like it took over South by Southwest and then they got the early adopters and then it became what it is today and and now it has hundreds of millions of users, right?
42:29
And then Foursquare did the same thing and there were few there were there were a few other uh companies that were able to do it and then after a while the the the kind of like the the area of opportunity and like the the shape that became more narrow and more narrow.
42:46
So like Foursquare made a lot of sense to launch at South by Southwest.
42:49
I don't know if they actually did, but that type of thing because it's like I'm checking in at this conf at this uh this uh stage for this show or this bar for this party.
42:59
Um then then they became more uh anonymous and there were a couple other apps and then the anonymous apps would go viral on college campuses and they would inevitably be uh kind of used for cyber bullying of one kind or another.
43:14
And so, um, always been always been a mess, but, uh, at the very least, you got to lock down the user data.
43:19
There was a funny funny post on Hacker News.
43:20
I was reading about this where someone was like, I want to set up a an app for, uh, people who enjoy doxing other people.
43:30
And so, it's a place where you can go to congregate with other people who enjoy doxing. >> The doc community.
43:35
>> The Yeah, the docs community.
43:35
So, if you're into doxing, you can go there and like connect with other people who enjoy doxing.
43:39
But when you sign up, you're instantly doxed and you show up on a different website that says, "I doxed myself." And so it's like, "Hey." Yeah.
43:48
We just need your we just need your address, your phone number, your ID, your image, your selfie, a bunch of photos, and then it just goes immediately public.
43:55
And so it's like a doxing honeypot. It's very silly.
43:58
So he's like, I was going to vibe code this, but I don't have time, so somebody else can do it.
44:03
Uh, anyway, DHH has some good analysis.
44:05
the founder of Base Camp, the legendary programmer, the creator of Ruby on Rails, >> the um we just we just got a text from a friend of the show about a friend of the show um >> uh and uh apparently um >> someone uh this person wanted to build a Yelp for people, >> which is actually which is actually like >> that's basically what tea app is, right?
44:31
Um, it's just like very one-sided, has this like dating dating market focus, but it would be really funny if if if there was just a public database where people could review John Kugan >> and just say like I don't like that he hypemoged me at that conference.
44:44
He knew he knew exactly what he was doing. >> Yeah.
44:48
I mean, like there's something weird about like anything that's synonymous like this, it just kind of goes to this like negative world, this negative realm.
44:54
Um, interestingly, Nikita Beer's app, Gas, even though it's not around or I don't think it's thriving anymore, but I thought it was interesting that that he you could not do free text response.
45:07
So, I couldn't say Jordy looks bad in a white suit or something like that.
45:13
I could only choose from four positive things.
45:17
And so, it was like enforced positivity.
45:20
And then I could be relative.
45:21
I could say like, well, he's more, you know, intelligent than charismatic or more charismatic than intelligent.
45:27
And I could kind of pick from different positive traits, which gave you a relative landscape of compliments.
45:33
But it but it it kind of it kind of prevented that like cyber bullying because you could only say nice things.
45:39
You could just say different nice things and then I think I I think that's what led to it going viral and and doing very well.
45:44
Uh and so uh there clearly are ways to work around some of these odd edges and we see them pop up in other in other ways.
45:50
We talked about it a lot with like the LLM stuff with like, you know, stuff comes out like social networking exists, the app store exists, the ability to import photos and stuff exists.
46:00
Like how do you use that stuff?
46:02
Like technology can be good, it can be bad, it can be misused.
46:05
When does when does >> Yeah.
46:08
When when stuff goes anonymous, it gets pretty dark and negative quickly. Yep. >> Right. >> Yeah. Yeah. it.
46:14
There's something There's something about not having >> That's why like if somebody's if somebody's talking Yeah.
46:19
badly about a certain investor online and not willing to like show their face. >> Yep.
46:25
>> You have to discount it massively because it's like, okay, was this founder like >> rejected by that firm once and now they have they feel like they were wronged and like their entire worldview is like >> oriented around that. >> Yeah. Yeah. It gets very odd.
46:39
Anyway, let's finish with the T app because there's an interesting conspiracy theory we got to go through.
46:44
We got to put on the tin foil hat.
46:45
So, first off, DHH, uh, founder of Ruby on Rails, founder of Base Camp says, uh, the T app having all its user data leaked was bad.
46:52
Now, imagine a porn site with an age verification tying viewing history to a specific identity.
46:58
I'm sure no hacker or government agency would ever have interest in such data.
47:01
So, he is uh very worried about that.
47:04
And then he also says, "Web app users would be shocked to learn that 99% of the time deleting your data just sets a flag in the database and then it lives there forever until it's hacked or subpoenaed."
47:14
This is because a lot of uh a lot of users demand this like, "Oh, I I didn't mean to delete it. Can you unddelete it?"
47:20
And they're like, "Yeah, we can.
47:21
Actually, we didn't delete it.
47:23
We didn't fully delete it."
47:24
>> Well, this this so um totally totally random, but over the weekend, I noticed so so in Sam's interview with Theo Vaughn. Yes.
47:32
for Theo's interview of Sam, >> he just kind of casually mentioned, he's like, "Oh, yeah, by the way, >> I've been thinking about this guy."
47:39
>> He's like, "By the way, like every single thing that you say >> on chat GPT can be used against you in a court of law and there's nothing we can do about it."
47:47
And he said like, "We need new laws." Yeah.
47:50
>> But it felt like something that the kind of thing that like he should probably have been like lobbying to create those new laws. >> Completely disagree.
47:57
Uh, >> so so so yes, maybe like public warning like, "Hey, by the way, anything you say to Chad GBT can be used against you in a court of law." >> Yeah.
48:08
>> Could could be done could be done at the app could be done at the app layer. Yes.
48:12
>> Do you think that that um >> and and you think it's like going and just yelling about it, you know?
48:16
So, so basically I think that that was a sort of like a PR blunder in the sense that what he should have said is like, hey, chat GPT, exact same rules as Facebook and Google Sheets and email and everything else.
48:32
Nothing's different about it. It's all the same.
48:34
Instead, he didn't make it clear that he's under the same rules as everything else.
48:40
And it would have been so much easier to to start there with that.
48:44
Hey, you know, we're a website that stores what you talk the the the text that you type into that box goes into a database and that database can be subpoenaed like any other website.
48:57
>> Any other website you you like if if you commit a crime, they can go look in your bank account.
49:01
They can go look in your email.
49:03
Like this is something that is that is they're not unique.
49:05
And and the way it was phrased in that interaction, it felt like Chachi PT was uniquely subpoenaable. And it's not.
49:11
It's the same as everything else. >> Same as iMessage.
49:14
Yes, the same as everything else.
49:16
And so, and so yes, there are endto-end encrypted apps where if the FBI can't unlock your phone or the signal messages delete, they can't get access to them. That's that's one thing.
49:28
>> So, all the headlines now are personal conversations could be used against you.
49:32
>> And it's the exact same thing with Gmail and it's the exact same thing with with Facebook messages.
49:35
And it's the exact same thing with, you know, sto saving a TXT file on your Mac.
49:39
Like like imagine if Tim Cook came out and was on the Theo Vaughn show and was like, "Oh yeah, like uh it's really interesting, but like all the files in your computer like the the government could like that could be evidence."
49:53
And it's like, "Yeah, obviously."
49:54
Or like, you know, anyone, we talked about this with like the papers in here, if there if the SWAT team comes through, they can take these like they could legally they can take this.
50:03
They can take this newspaper and see, oh, what was John reading?
50:05
Was he reading about crimes?
50:07
reading about crimes? was what they can link this maybe I was circling like oh do crime like you know maybe there's a maybe there's an article in here about how to do crime and I was reading that like that would be admissible in the court of law it's no different and he didn't >> friend of the show says iMessage can't
50:23
be subpoenaed if you have them on autodelete Apple doesn't store them >> yes yes yes that's true and and iMessage is end to end encrypted in a way that iMessage doesn't uh Apple doesn't have the server Apple doesn't have uh the messages stored on their server um this is the same for WhatsApp app with encryption. But again, it's it's not a
50:38
But again, it's it's not a uh like like I don't think the expectation should have ever been that chatbt was end to end encrypted because they've never said that.
50:49
Why would that be the expectation?
50:51
I I mean I get that people don't understand how inference happens, but you hear oh my my everyone's talking about the water, you know, usage and the energy usage.
51:00
It's like where do you think your prompt is going?
51:04
>> Do you think it's happening on device securely?
51:06
securely? Like do you think that's that's what I understand people don't people don't get it like it is complicated like we we we we talk to Ben Thompson I mean analysis all day long we understand this stuff >> but he could have done a lot better I
51:17
mean it would be hard to avoid the headline in general of saying your chat GBT conversations can be used against you because somebody could extrapolate that >> from even if you just said all consumer tech can be can be subpoenaed. >> Yes. And so and so on the law side, what >> Yes.
51:33
>> Yes. And so and so on the law side, what I would advocate for is if OpenAI wants to release something that is like a tiny box or a device or a Johnny IV device that does the inference locally or does it in a secure encrypted end to end with deletion like like methodology basically
51:54
like in like like it's secure that would be a great product and people might want that and that's what George Hott is advocating for >> or a specific therapy We've even seen >> a a a like you know a specific product for therapy, a specific product for legal work. >> Oh, totally. Totally. There's >> Oh, totally. Totally.
52:09
There's >> Yeah, but right now there's no way to do that unless I bet you if you figured out how to run the inference some way in an email chain with your lawyer, like you could actually make that attorney client privileged.
52:27
I don't know exactly how that would work, but I'm pretty sure if like imagine imagine if every chat GPT uh request I email my lawyer and I say I need you to prompt this and then send it back to me.
52:39
That would be attorney attorney client privileged.
52:41
And and if I did that for my like a therapist as well, if I went to a therapist and I said like here's the prompt that I want you to run and then send it back to me, that would probably be like privileged in a medical context.
52:54
But the idea that Chachi BT just out of the box would be would be privileged in some way should not he should not have set the expectations there when he when he talked about that.
53:03
Um the the the flip side is of course um uh yeah the this ondevice there's even that meme that we were talking about the the if you didn't if if your AI girlfriend isn't running locally that's not your girlfriend right like like that's a good way to to to relay that information.
53:23
Um, but OpenAI doesn't have a product for that yet.
53:25
And so, uh, it it was kind of an odd way to to to noodle that out.
53:30
Um, but I think people will learn and I think people will will, uh, understanding this.
53:34
Uh, anyway, uh, let me tell you about linear.
53:38
Linear is a purpose-built tool for planning and building products, meet the system for modern software development, streamline issues, projects, and product road maps.
53:44
You can start building at linear. app.
53:45
Uh DHH was talking about how he implemented how he built uh hay and base camp.
53:51
Um especially when it comes to deleting log files, database backups, and other ants auxiliary copies of your stuff most companies just hang on to until the sun burns out.
54:03
>> What if what if DHH built this so well and so secure that it's just the most loved email provider in for narco terrorism?
54:12
>> That's always the risk. Yeah.
54:12
I mean uh you you talked to Moxy Marlin Spike who built Signal.
54:19
>> He says content can't be recovered once it has been permanently deleted. >> Yeah.
54:23
And so yeah, it really is massive effort enterprising.
54:26
>> It's not just a flag on the database because that data also goes to the log files also goes in the database backup.
54:32
So let's say that you imagine that you're like yeah of course we back up our data to another another cloud storage provider.
54:38
Uh, we serve most of our users on AWS, but just in case AWS goes out, every month we archive all of our data and we send it over to Google Cloud Platform, GCP.
54:47
It's like, that would be amazing. I wouldn't lose my data.
54:51
But now, if I want to delete my data, I have to delete it in two places.
54:54
Well, what if you're exporting it to tape drives and putting in Iron Mountain?
54:57
Like, this is a thing companies do.
54:58
Like of course like I want to have triple backups but now you have to go delete my data from AWS and then from GCP and then how how like you know all the different backup methodologies >> including uh there has to be some real moes involved.
55:14
People usually talk about moes in the context of you know >> strategic uh but like if you had an a moat full of alligators >> defense >> securing you know a handful of you know large hard drives. >> Yeah.
55:28
Could could be something there.
55:29
>> We got to find a We got to find a house with a moat. I'm sure it's out there. It's going to come up.
55:33
I feel like the mansion section of the next J.
55:36
>> Our next studio when we do the ground up build >> get a big enough property to have a moat. >> Definitely.
55:42
>> It's so hard to get a moat these days in business.
55:43
One of the easiest ways is to just physically build one.
55:46
>> Everyone needs a moat.
55:46
It just sets you up for the mindset of having a moat.
55:50
>> Um anyway, we got to put on the tin foil hat for this one.
55:53
There's something really weird >> because I don't I don't uh I don't buy this theory. >> You don't buy this.
55:57
There is something really weird about the TAC geocode data and I'm having trouble putting my finger on it.
56:02
Dirty Texas hedge says, "I hate making speculative accusations, but the most >> which is the name hedge dirty." >> Yeah, weird.
56:11
>> Sounds a little >> I hate posting wild conspiracy theories, but here goes.
56:14
Um, but the most explor explanatory theory for this data is investor fraud of a particularly ingenious form.
56:22
Again, this is a speculative theory, so take it as such.
56:26
First, the most likely explanation for the geographic pattern described is that the data is fake and randomly generated.
56:33
If you want to, it's not it's now not particularly difficult to mass generate realistic photos with whatever metadata you want.
56:40
The other reason to suspect this data is fake is if tens of thousands of women had data exposed, the internet would be a wash in horror stories of incel trolls trying to ruin their lives and crickets.
56:49
And I feel like there's something there where if you were in the 75,000 that apparently got leaked and you put a and you put a Tik Tok out that says I was one of the people that was hacked, it's like it only has to be one out of the seven 75K that comes forward and says like I'm I'm affected. I'm being trolled. Like support me. Here's my GoFundMe.
57:08
Help me pay for a new social security. >> Yeah.
57:12
But all these people speculating, have they talked to any of the women affected?
57:15
It's it's saying it's also wasn't Yeah, it wasn't you know the >> um >> the IDs are real.
57:24
>> We know that >> the I mean >> I I >> So you're saying you're saying uh I saw images of the IDs.
57:31
>> You can just generate a fake you can generate a fake real ID or like a fake a fake image of an ID that looks real.
57:35
How would you actually understand what the theory here?
57:39
They were buying app downloads to get to the top of the chart and then they they generated a bunch of fake and they had a fake data leak.
57:47
>> So this happened with Carly Javvis. You remember this? >> Yeah.
57:50
So this you're talking about Frank.
57:51
So Patrick McKenzie's in the comments here.
57:53
I remember Frank selling one of the world's largest financial institutions a list of users who were not actually users and in many cases didn't actually exist.
57:59
So, this was a company that was like a like Gen Z finance bank uh platform sold to I think JP Morgan >> and then JP Morgan realized postacquisition that it wasn't like the the users weren't real people.
58:12
Um but I don't understand >> in here and he and he he is confirming that some of the ideas are real because you can Google the names and they don't seem hallucinated by an LLM.
58:22
And also if this is from 20 2024 the AI image generation technology was much worse back then.
58:28
So, I think there would be more tells of like you look in the background and you see some sort of like nonsense. >> Yeah.
58:34
I just don't understand like you have to argue here that like they bought a bunch of bot farm downloads and that's why they're charting and they're not actually at the top of the charts and they were and then he he did a fake data leak. >> Yeah.
58:48
>> To try >> Jeff Lewis Jeff Londale is in here and has a funny one.
58:50
He says so they were sloppy with some fake geocoding but dead on in their generation of the images themselves. >> Yeah.
58:56
I think I think the argument for the geocoding data is like they possibly set it up in a way where they just said where what city the person is in because even on the map that they show it doesn't give it just puts a pin on the local town.
59:08
>> So So the problem with the uh w with that theory is that if you look at the map it's it's like it's distributed like almost based on geographic space as opposed to population space.
59:21
So you would expect like Manhattan to be flooded and you would expect Los Angeles to be flooded in Chicago, like the big cities.
59:28
But instead, if you trace the the lines, you'll just have like a random user.
59:33
You'll have like one user in you'll have one user in Los Angeles and then like one one user like on the five every five miles all the way up to all the way up to San Francisco.
59:44
And you know that from driving through if you drive from LA to San Francisco, it's like LA is massive, San Francisco is massive.
59:51
there's some stuff in Bakersfield and then it's like four hours of nothingness basically in like farmland.
59:56
And so the idea that you'd have just as many users out in the middle of California as opposed to in the in the in like the the like the hot spots uh or like the the the population dense places is weird.
1:00:07
But there are maybe examples for that where maybe the data was filtered and the data that was leaked was was sort of like some sort of distribution based on location. Exactly.
1:00:22
But ideally, in theory, if you did random sampling on basically any app with 60 million users, you would see major major clusters around big population centers. So, it was a little odd.
1:00:31
Um, so, uh, an app that with fake data is most likely trying to default to, but they didn't really raise that much.
1:00:39
So, it's kind of like who did they do this for? Maybe they did this.
1:00:42
>> I don't know that they raised at all. >> Yeah.
1:00:43
And so, yeah, I mean, maybe there's the maybe the theory is like they they do this fake data.
1:00:46
They try and go out and raise for it.
1:00:48
they couldn't raise, but then they left the fake data up and then the fake data gets leaked.
1:00:52
But the weird thing is that like why would you leave your fake data like dump that you generated like still accessible and hackable? Like that's weird.
1:01:00
And so you have to really get into like 40 chess mode.
1:01:02
So I'm giving this tinfoil hat conspiracy >> uh two tinfoil hats out of five. >> Two out of five. >> Two out of five.
1:01:11
>> That's a lot of hats, John. >> 40%. That's a failure.
1:01:15
>> What would you give it? You give it zero. You don't buy any of it.
1:01:16
I don't >> I'm still I'm still a little weirded out the geo code.
1:01:21
>> Yeah, the geocode data is weird, but there's so many other explanations other than like you have to make so many leaps from that point. >> Yeah. Yeah. Yeah. Yeah.
1:01:28
Well, if you have users that are in a whole bunch of different geo codes, you got to pay sales tax.
1:01:33
You got to get on numeral hq. com. Sales tax on autopilot.
1:01:35
Spend less than five minutes per month on sales tax compliance at numeralhq. com.
1:01:42
Mark Cuban and you got in a fight, a knockout dragout fight.
1:01:44
The timeline was in turmoil.
1:01:48
>> Mark, if you're listening, you're welcome on the show.
1:01:49
We'd love to debate this with you in person, but I will be playing the steel man for this debate.
1:01:52
I will be arguing in favor of banning AI ads.
1:01:58
So, we've been in AI ad turmoil before because we had uh the CEO of Plexity on the show, asked him said, "Hey, we love ads.
1:02:07
We love ads on this show.
1:02:09
We would love for you to run ads. We're all pro ads here."
1:02:12
He said, "Yeah, maybe we'll run some ads."
1:02:14
And Techrunch took it out of context and said, "Perplexity is going to put ads everywhere."
1:02:19
Uh, which maybe they should, but clearly people have a visceral reaction to ads.
1:02:22
The classic stated preference versus revealed preference.
1:02:27
Everyone says they hate products that have ads in them.
1:02:29
And in fact, they use products that have ads in them all the time.
1:02:33
>> Can you call that one one example where we are both happy to pay for X to not have ads? >> Yes. >> In X. Yes.
1:02:39
But my one one reason that that I feel that way is that X ex X programmatic ads were just never that great.
1:02:48
You I I cannot remember a single time that I purchased a product because I got it >> uh uh got an ad for it on X. Yeah.
1:02:55
>> So this all started >> really quickly.
1:02:57
What's really weird about the X ads is that I feel like if I could turn back the clock and I was in charge of like good X ads, like I would have just been seeing is this a straw man hat?
1:03:09
Why do I have the straw man hat on?
1:03:11
>> Because you're you're siding with >> Cuban. >> Cuban.
1:03:14
>> No, that's the steel man. I'm steel maning.
1:03:16
>> Oh, you're steel maning.
1:03:16
>> So, I will be putting on the steel man.
1:03:18
The straw man is for an argument that no one is making. >> Okay.
1:03:21
>> Uh it's because it's a fake fake argument. >> We'll get the steel. Get the steel. >> We'll get the steel.
1:03:24
But first, I want to talk about X ads.
1:03:25
I feel like if if if X ads were going to be great or or had become great, um it would have looked like our advertiser lineup.
1:03:32
I should have been scrolling my feed and seeing like ramp, linear, graphite, Figma, Vanta, you know, these these companies should have been the ones because that's like like the enterprise SAS buyer is absolutely on was on Twitter and still is on X and yet for some reason those ads it feels like the natural place.
1:03:56
feels even better to advertise a an enterprise SAS product on X to Teapot as opposed to or just tech Twitter broadly as opposed to on Instagram.
1:04:06
Um because you're right there.
1:04:08
Yes, it's going to be very low conversion because you're mobile.
1:04:10
You're just kind of scrolling.
1:04:12
You're looking at random text.
1:04:13
You're not in enterprise buying.
1:04:15
In terms of just building name recognition, it feels like that should have been like a growing category of ads, but instead the ads were always these like very low tier like sloop teu level products. Yeah.
1:04:28
I saw a lot of Lee's like it's a projector that that shines like a universe on your on your like have you seen these? Yeah.
1:04:35
It's like a projector that shines a universe on your uh >> simulating a sunrise. >> Yeah.
1:04:41
On your uh on your ceiling.
1:04:41
It was like it was like completely backtory. >> July 26. >> Okay.
1:04:48
>> Mark Cuban hits the timeline. >> 127 on a Saturday. Hey, David Saxs.
1:04:54
>> My one request is that we make it >> you. You missed the space. Hey, David Sax.
1:04:58
Space, >> my one request is that we make it illegal for AI models to offer advertising.
1:05:04
And we need to really exam examine referral fees as well.
1:05:07
The last thing we need is to have algorithms designed to maximize revenue driving LLM output and interactions.
1:05:15
I would just say here every for-profit company is set up to maximize revenue regardless of how they like that is the goal of a I guess I guess you could argue that open ad's a nonprofit.
1:05:29
Um but he says they are already recommending brands and we don't know if they're getting paid for it.
1:05:33
We need to have our We need to have learned our lessons from alos in social media.
1:05:40
Pull pull out the helmet. >> Is so intense.
1:05:42
It has a >> I'm not going to be able to. >> It has a lock on it. >> That's great. That's great.
1:05:46
Um and so I started by saying guy who made his money selling ads online wants to ban selling ads online. >> There we go. >> John is locked in.
1:06:01
I need to strap it in fully this time because last time it was shaking around a lot and I had to >> Sean Sean Tanu in the chat says, "Dude, I'm in a test flight for an adbased AI chat router mobile app for free reasoning will ship by next week." >> Interesting. >> So that's cool. They're coming. They're coming. >> Okay.
1:06:22
So the argument for banning AI >> Can you put it on fully, please?
1:06:26
>> Yeah, you got to put it on full. There we go.
1:06:28
>> The argu Can you hear me? Okay.
1:06:29
>> Yeah, I can hear Okay.
1:06:29
The argument for for banning uh advertising in AI chatbased models is the same argument for banning not just Tik Tok but all short form video, all brain rot apps, all slop apps.
1:06:44
all slop apps. So many people when they make the argument that Tik Tok should be banned, they make it based on geopolitical considerations between China and America and they say that Tik Tok is spyw wear or Tik Tok could be manipulated by the CCP to change
1:07:02
political preferences in the United States or potentially um do something harmful to the American population, get them less focused on math and basically creating shareholder value and instead more focused on just uh you know arguing with each other about whatever is viral that day. And there's this famous
1:07:19
And there's this famous example of like oh well like the Tik Tok in China just shows you math and education videos and the Tik Tok in the US shows you like you know random you know slop stuff and controversial videos to get you to never stop watching.
1:07:36
And so you have to put all that aside because this is not a geopolitical discussion.
1:07:41
This is a discussion of business models, but there are people too who argue that it's not just Tik Tok and you should go farther and you should in fact ban YouTube shorts and Instagram reels and >> ban video >> and the reason >> and video if it's short and vertical. >> Yes.
1:07:59
And the reason is that um >> it's a it's a terrible drug for the mind.
1:08:04
It is addictive and it leads people to uh drop out and stop focusing on you know longer more thoughtful things.
1:08:12
We need to go back to reading books.
1:08:13
We need to go back to watching films something you're not familiar with.
1:08:18
You need we we we need to appreciate the arts.
1:08:21
We need to appreciate uh craft.
1:08:22
And there's no craft in a 60-second vertical video.
1:08:25
And therefore, similarly, how did we get to short form vertical video?
1:08:33
these endless timelines, these endless scrolls.
1:08:35
We got there through advertising.
1:08:36
We got there because the longer that I can keep you on the app, the more ads I can show you, the more money I can make.
1:08:44
So, it's this natural economic impulse.
1:08:46
And so, if we do this with LLMs, the models will no longer be optimizing against giving you the most concise answer, allowing you to move on with your day.
1:08:58
They will be baiting you into endlessly chatting with them all day long.
1:09:03
Take you down some crazy rabbit hole.
1:09:05
You'll just have asked, you know, something basic like, I don't know, like how do I tie my shoes?
1:09:11
And then all of a sudden, it's like giving you the history of shoes and taking you over here and telling you about the controversial nature of certain shoes and blah blah blah.
1:09:20
And you'll just be sucked into this.
1:09:21
You'll you you'll you'll tune out of everyday life.
1:09:23
You won't be talking to your friends.
1:09:25
You won't be talking to your loved ones.
1:09:27
And you'll become obsessed with your phone.
1:09:32
You will be obsessed with your LLM.
1:09:32
And you will ultimately be be brained >> on steroids.
1:09:38
To George Hotz's point, imagine a future where you have 10 CIA agents >> tracing all the time, convincing you to buy things >> at all times.
1:09:51
>> And so that so yeah, so uh relevant post here from Rune.
1:09:54
Obviously, he's conflicted.
1:09:56
I I you have to imagine Fiji Simo will roll out an ads product at some point at OpenAI.
1:10:05
>> Run says, "Advertising is far more aligned business model than many others.
1:10:08
It has been vilified for years for no good reason.
1:10:11
>> User minutes maximizing addiction slop would exist with or without it." >> Yes.
1:10:15
>> And the reason for that is like you could have a subscription based product and there's still an incentive for the company to try to get you to use the product more than any other product.
1:10:24
so that you keep your subscription. >> Yes.
1:10:27
>> And you upgrade to higher tiers and things like that. >> Read his example.
1:10:31
>> He says Netflix CEO with subscription pricing only on record saying we're competing with sleep. >> It's true.
1:10:39
>> And uh and I think Netflix they do have an ad supported model or they've considered it.
1:10:43
They've considered rolling it out but that was at a time when he said we're competing with sleep.
1:10:49
Run says often times when going on Instagram the ads are more immediately high utility than the reels.
1:10:52
It's pretty incredible when you can monetize the user in a way that actually values adds value to their life.
1:10:58
This opinion is basically a relic of the late 2010s consensus that Facebook is an evil company, but has more to do with them than advertising generally.
1:11:07
Um uh and so yeah, he says people misattribute this incentive problem to ads when it's native to all webscale products.
1:11:14
And so yeah, my my general take is that um uh AI is simultaneously very similar to other consumer tech products >> in the sense that uh ads have been the primary economic engine of the internet to make products free or cheaper and make digital products and services widely widely available to all. Right?
1:11:42
And in some ways AI might be different but at least in the short term I think it will be very similar.
1:11:45
Um today uh open AAI has had an extreme incentive to drive paid subscriptions to the product.
1:11:55
So in that way they want to make the product the free product valuable, useful, maybe make it even addictive, right?
1:12:01
Maybe they want you to get addicted to talking to your, you know, chat GPT like a therapist so they can upsell you on on um, you know, so so make it so you're less rate limited, etc.
1:12:12
And so my my issue my primary issue with with Cuban was saying blanket ban on all ads in AI.
1:12:18
I mean, that would just be insane.
1:12:21
And I think it would naturally lead to people without the ability to pay for subscriptions having less access to highquality tutoring, the infinite knowledge engine, etc. , etc. >> Mhm.
1:12:34
>> And so basically saying like we're going to just completely ban the the the the thing that made content on the internet and services and apps free.
1:12:42
We're just going to completely ban that thing.
1:12:45
um that felt like just way too extreme.
1:12:47
Now the concerns that he has around okay we don't know if these companies are being paid uh we don't know if the the he said we don't know if if Chat GBT is getting paid to promote certain products more than others and I would just go say anecdotally I don't know a single company and that is paying a foundation model company directly >> to directly or indirectly >> and I think the FTC already has rules on So, so this has been solved in two ways.
1:13:18
In search, >> you know, when you're searching, you know, is this something that is been paid?
1:13:23
Is this a paid placement? Right?
1:13:26
You can see that it says ad >> and uh and then you can see that the SEO, >> the UI around whether it's an ad or it's not has changed a lot over the years.
1:13:36
And Google used to put like a big yellow box around it and make it very clear this one was an ad.
1:13:40
this one was an ad. that has been all dialed back over the years very slowly and the hey this is an ad little note has gotten smaller and smaller until it is somewhat indistinguishable but I agree with you and I would be very surprised if openai was running ads without disclosing them that would just
1:13:57
be an insane L and it would be it would open up so much liability that would be crazy >> yeah and and knowing enough various startups that would love to pay >> every company would love to to have the the super intelligence or AGI say actually I know you searched this company but I really think you should consider this one too. >> Totally. Totally. >> Totally. Totally. >> One thing is true.
1:14:16
So companies like profound help with effectively AI SEO like AI observability and they help you do things that improve your visibility. Yes.
1:14:27
>> In models and that is the exact same thing that companies have done >> from an AI ad ban. Correct. >> Correct. >> Yeah. So we're team profound.
1:14:36
So uh >> maybe that's the bullcase here.
1:14:40
So, so yeah, suddenly suddenly >> funnel all the money into AI SEO >> and so so but but the other concern here is like if you basically say like uh companies need to give away models completely for free or they have to charge money >> that that is not letting the free market do the work of just saying like we should have everything.
1:15:00
You should have ad supported models.
1:15:02
You should have >> free open source models that you can run locally.
1:15:07
You should have paid super intelligence, the best model, no rate limits, etc.
1:15:12
And the other thing here is like >> so many searches and so many of the ways that people get value out of models are just completely non-economic, right?
1:15:22
They have zero purchase intent. >> Okay. >> What?
1:15:27
>> But the counterpoint to I agree with you.
1:15:31
If I go and I search for, you know, best insurance and best car insurance in Los Angeles, Google's going to give me a ton of ads because that's highly monetizable.
1:15:39
If I go and I go, >> and it'll also give you LLMs or sorry, it'll give you SEO based results. Yes. Yes.
1:15:45
>> That are heavily optimized.
1:15:45
People have >> paid and then and then on the other side, if I go and I ask, you know, like what what day does Thanksgiving fall on this year?
1:15:54
That's not heavily monetizable.
1:15:56
And so Google will just give me the answer without a lot of ads, right?
1:15:59
But in the LLM example, if I ask for something that's nonmonetizable, I just ask, hey, you know, what's, you know, 70 75 time 164, >> the LLM actually does have the opportunity to try and take me down a more monetizable path. >> Yeah.
1:16:17
And so that could be annoying, but it could also be and it could cause churn if it's not done well, but it also could just say, okay, I've I've delivered this user, you know, the what they asked for, but now my goal is to keep the app open and show them ads.
1:16:30
And so let me What day is Valentine's Day?
1:16:35
And then it's like, oh, like you kind of botched Valentine's Day last year. Remember? >> Exactly.
1:16:40
>> Remember we we we were talking about how uh your your girlfriend wasn't super happy.
1:16:45
I think this year hit her with flowers, hit her with concert tickets.
1:16:49
>> Now, the problem is if it did that, that's actually a helpful ad and that's actually good, right?
1:16:52
Like if you did in fact botch Valentine's Day last year and the LLM and you're just asking randomly and then it's like, hey, I'll help you and I can order flowers from this place because they bought an ad.
1:17:01
Like that's actually consumer value. I I >> Yeah.
1:17:04
And it's like, show me four places that that that uh show me four places that and and then the question is like if you ask the model Yes.
1:17:12
Is it only give you paid results? >> Yes.
1:17:16
>> And if it does, it should >> it should >> it should have to disclose that, right?
1:17:23
We've solved this this in the in the influencer space.
1:17:24
Influencers if they're being paid specifically to post about a certain product. >> Yeah.
1:17:30
>> They need to disclose that that um that they that they were paid to do so.
1:17:35
>> Like if we were being paid to advertise for Adio, we would have to tell you that it's customer relationship magic.
1:17:38
Adio is the AI native CRM that builds, scales and grows your company to the next level and you can get started for free and thank you to Adio for making this possible for sponsoring this show. So the last Yeah.
1:17:53
So I would say like my general take was like blanket ban is bad probably.
1:17:57
Completely agree with that. I can't steal man.
1:17:59
>> Probably need some new guidelines and new even Sam was saying we need laws around what can be subpoenaed, what what's private, what's not.
1:18:06
So probably need some new laws.
1:18:08
It's probably going to follow pretty closely what we already kind of what happened with search. Yep.
1:18:15
>> What happened with influencer, what happened with with all these, you know, different um categories.
1:18:19
But like fundamentally, every AI company has an incentive to get users to use and trust the apps more than other apps. >> Yep.
1:18:32
>> And an incentive to monetize them because most of the time they're for-profit companies that have a profit incentive. Yep.
1:18:40
>> And uh and I think that it's something that everybody should be widely aware of, but at the moment I'm not aware of examples of this being this sort of like trust being abused.
1:18:51
And I think that ads have the potential to make to make it so that uh kids have a tutor in their pocket or a therapist in their pocket.
1:19:01
And if you and if you take away that economic engine, uh it could lead to some uh bad outcomes.
1:19:09
>> Yeah, I'm ready to take the steelman helmet off.
1:19:10
Mark Cuban will have to finish this one on his own.
1:19:12
It's really funny to think like, okay, you could have social media and brain rot for free, >> but if you want a tutor that is going to help you excel in life or you want a therapist who can be there for you, you're going to have to pay for that buddy or you're going to have to take the the the the free, you're going to have to run that that uh open source model locally, buddy. >> Yeah.
1:19:34
I I I think the big the big question is just like how much how how much will ads change what's already happening in chat GPT as a product?
1:19:45
Because if I'm a product manager and I'm trying or the CEO and I'm trying to grow revenue, I would imagine that I'm trying to make the free tier as engaging as possible so that people upgrade to the paid tier.
1:19:57
And and I don't know that my incentives change dramatically with ads versus just paid upgrades because I'm trying you're you're on the free tier chatbt.
1:20:08
I'm trying to keep you engaged and and satisfy you.
1:20:11
And so the incentive towards like user satisfaction, maximizing user minutes.
1:20:18
It feels like it exists in both the ads.
1:20:21
>> The final state, John, you know what the final state is?
1:20:23
UBI, like universal basic tokens.
1:20:26
And and >> now I was gonna say the final state is you pay a lot of money for intelligence and you still get ads because even like think about it.
1:20:36
If if Google was if Google was charging if Google was charging you like $5,000 a year for search, I I bet you they make more than five grand from you a year in search >> from from the search product, right? >> Yeah.
1:20:50
And so they actually have an incentive to say search is free, John. >> Yeah.
1:20:54
And didn't >> search your search your search your little heart out. >> Have ads. >> Yes. Duck. Go has ads.
1:21:00
Um I was wondering if uh so they are not based on tracking users building personal profiles.
1:21:07
So there was clearly some sort of backlash to Google.
1:21:09
It was a stated preference revealed preference thing where most people are fine with Google ads. But Duck Duck.
1:21:15
Go did grow as a business and become a big business, but they still have ads.
1:21:20
And so now they they they don't track users and they they put privacy, you know, more uh >> well, the critique the critique of big big tech broadly was they sell your data, right?
1:21:32
Like they they're collecting all this data on you and they're selling it.
1:21:35
It's like no, they use the data they collect on you to deliver you hyperpersonalized ads. >> Yeah.
1:21:41
If I'm if I'm advertising on Facebook, don't give me data. Give me customers. You take the data, man.
1:21:47
>> Yeah, that's the >> You take the data. Don't give me the data.
1:21:49
I don't know what to do with the data. >> Distribution. That's That's valuable.
1:21:52
>> You You have the AI scientists making a hundred million dollars a year.
1:21:54
Use them to to, you know, get clicks on my website. Do not send me the data. I don't care.
1:22:01
Um, but I do care about Finn AAI, the number one AI agent for customer service, number one in performance benchmarks, number one in competitive bakeoffs, number one ranking on G2.
1:22:10
You can start a free trial at Finn. ai. AI.
1:22:12
Um, >> what else we got, John?
1:22:16
>> Uh, how the EU succumbed to Trump's >> Oh, we got we got to go through a couple of these.
1:22:20
So, I just wanted to add some color.
1:22:22
So, Antonio Garcia Martinez, who's coming on the show tomorrow, right?
1:22:27
>> He says, >> "I would bet my entire net worth that we will have ads and AI.
1:22:30
If they take the form of highly relevant offers, users will welcome them.
1:22:35
They'll have a sky-high conversion rate and web shopping will die.
1:22:39
and they will be necessary to pay for the compute in many consumer AI apps.
1:22:44
And this third part is what I was saying, right?
1:22:46
Like you either have if you want this incredible tutor in your pocket with every student.
1:22:52
Some students will be able to pay, others won't be.
1:22:53
And if you can create a great ad engine, uh that that you know, people will be able to have access to these tools that otherwise wouldn't.
1:23:00
And so this is a pretty crazy parlay uh AGM crazy parlay for your entire net worth.
1:23:07
But uh I think I'm riding with you.
1:23:07
Go express it on poly market.
1:23:11
>> Yeah, we gotta get a we should get a we should get a poly market.
1:23:15
>> Was in a similar similar situation too where uh you know prediction markets were a new thing and the government needed to figure out how to regulate them.
1:23:23
That wound up happening and now they have uh approval to >> work Michael McNano says agree with all of these and I'll add one ads and AI will be the best money printing machine in history better than Google and Meta.
1:23:36
The reason that the reason that I think this will probably be true is that the you know one of the number one ways people decide what products to build >> is through recommendations from friends and I believe that these the you know
1:23:52
chat GBT and and other players will effectively serve that role of like open AI will have to manage the like user trust right because if you like get an ad for a product on Instagram and it's just okay you're not mad at Instagram. If you get an ad on Instagram and the
1:24:07
If you get an ad on Instagram and the product comes and it's like not at all what was advertised, you actually have Meta will just kick you off the ad platform if you do it like a couple times. >> Yeah. Yeah.
1:24:16
And they have a review system too.
1:24:18
After you buy something, Meta will actually ask you the challenge is like if you have uh chat GPT just like glazing some product.
1:24:25
Oh yeah, this is this is the best product ever.
1:24:28
It's not just a product, it's a lifestyle.
1:24:32
And then you get the product and it's mid.
1:24:33
mid. people are gonna people are not you know it will lose its effectiveness but if it's actually the sort of like trusted source for product recommen recommendations both organic and paid >> um yeah >> so anyways I would say uh credit to Mark Cuban because he eventually >> said >> blanket
1:24:54
but I think he's I think he's opening up a genuinely important conversation to have and I and I think you got to hand it >> someone else said if AI ads are clearly labeled and separate from real conversations is fine, but once they blend in with human chats, it's not marketing anymore. It's manipulation. It's manipulation.
1:25:08
And Mark says, which is the ultimate goal of most advertising manipulation, which is like Yeah. like >> companies. Yeah. Kind of loose.
1:25:16
>> You should just ban advertising and and see what happens to the economy. >> Yeah.
1:25:20
Um maybe he can make that part of his um he eventually says I think there will >> um he where he says where I could see AI ads being okay as if they're just listed as a chat and identified as an ad >> completely independent from user generated chats.
1:25:36
So I'm imagining he's saying like put it on the sidebar.
1:25:40
>> When have you bought a product from like a display?
1:25:42
He's basically >> promoting like display ads which >> still >> kind of a throwback >> people still run.
1:25:48
they're still valuable to to do, but um >> are there any areas where ads are truly banned?
1:25:54
I feel like certainly not at the federal level.
1:25:57
Like the FTC enforces or the FCC enforces a lot of advertising bans on certain products like like gambling and cigarette companies are banned from advertising.
1:26:08
But I it's it's odd to think about flipping it around and saying like this venue must be adree. Yeah.
1:26:16
It's usually a it's usually left to the free market and it's just a choice.
1:26:21
So So you know a a director can choose not to do product placement in their ads or in their in their movies.
1:26:27
A you know a TV show can choose not to you can host a podcast as adree allin would chose to be adree.
1:26:35
Yeah >> and that was just a choice that they made and that was a differentiator and some people like that and that that that confers certain benefits and costs.
1:26:42
you don't make the money from the ads, but maybe your audience likes that more.
1:26:46
You know, there's a whole bunch of different ways >> summed it up.
1:26:49
Uh, somebody responded to my post saying that uh banning ads would make the best AI inaccessible for lower income Americans.
1:26:59
And uh, Grock says, "Cuban's proposal ban ads to ban ads and AI models to curb revenue based outputs.
1:27:08
Without ad income, companies may rely on subscriptions pricing out lower income users from premium features.
1:27:12
This locks of SAI behind pay walls which could widen inequality.
1:27:16
The wealthy gain advanced tools for productivity and learning while others get inferior free versions entrenching economic divides over time.
1:27:25
>> So who knows if Grock is right, but um that was high level uh my concern with the proposal. Why don't you read?
1:27:33
>> Speaking of ads, here's an ad for eight sleep. Get a pod five.
1:27:34
5year warranty, 30 night risk-free trial, free returns, free shipping.
1:27:39
Uh, let me read climbing my way back to the top of the charts.
1:27:41
I got an 83 last night, eight hours and 24 minutes.
1:27:46
>> I'm gonna read through this Joe Weisenthal postp uh uh you'll be back in a second.
1:27:49
So Joe Weisenthal says Trump is winning on trade and uh Bloomberg opinion has a fantastic post here linking uh Trump's uh tariff brawl with Hulkamania.
1:28:00
Of course, Hulk Hogan passed away last week. Very sad news.
1:28:05
Um, but let's read through it from John Author's in Bloomberg opinion.
1:28:10
He says, "His trade victims never got their defense together to take on the bullying."
1:28:15
So, Hulkamania and global trade and markets in the week around the sad passing of pro- wrestling legend Hulk Hogan.
1:28:23
One of his most devoted fans, President Donald Trump, has been honoring his legacy.
1:28:27
Wrestling is a staged performance where the winners often portray themselves as bullies.
1:28:33
Trump is getting results from Hulkcommania in the much tougher world of international trade.
1:28:37
Sunday brought news of a trade deal in quotes with the European Union sealed at the president's golf course in Scotland which he described in a Hoganesque language as the greatest deal of all time.
1:28:49
Uh in it the EU accepts tariffs of only 15% on its exports to the United States and levies zero tariffs in return.
1:28:57
This has been largely expected as Japan as Japan's similar deal several days earlier left the Europeans little choice.
1:29:08
The EU agreed to buy $750 billion in energy from the United States to invest $600 billion in unspecified ways that it wouldn't previously have done and to buy American arms and weapons.
1:29:20
Uh the deal isn't a trade treaty.
1:29:23
Such things cannot be thrashed out in a 45minute meeting at the golf course.
1:29:27
It's barely even about trade.
1:29:29
and the EU gets nothing from it.
1:29:29
To use a phrase from tigress financial partners uh Jean Igrass uh Ergus, it's more the extraction of reparations from Europe's for perceived past wrongs.
1:29:42
And yet it's market friendly because the US had threatened to levy a tariff of 30% on EU imports from Friday.
1:29:50
In possibly the biggest victory for Trump, stock markets have brushed off the excitement to set all-time highs.
1:29:56
That is of course why we're wearing white suits today.
1:29:59
All-time highs in the markets.
1:29:59
The deals with Japan and the EU and other recent days follow massive concessions to the administration by the media group Paramount and Columbia University.
1:30:08
The classic Hulk tactics have worked and opponents have been picked off one by one.
1:30:14
Despite game theory to the contrary, which Points of Return covered back in April, bullying has paid off.
1:30:20
Game theorists show that bullies can be beaten if the victims stand together and take some pain.
1:30:25
The bully will hurt more than they do.
1:30:27
The rest of the world seemed ready for this a few months ago.
1:30:31
US trading partners from China to Canada and through to the EU immediately threatened retaliation, but now they're caving one after another. How has this happened? Facts have helped.
1:30:40
To date, tariffs have produced a lot of revenue for Washington without clear negative effects on inflation or US profits.
1:30:47
The economy is doing very well. It's very strong.
1:30:50
That strengthened Trump's hand and made him more credible.
1:30:53
Beyond that, the bully has convinced people he means business with renewed and escalating threats, and his targets haven't coordinated their defense.
1:31:01
To grasp what might happen next, let's look at the deal with Japan, another open market that depends on exports more than the US does.
1:31:09
Local stocks, also held back by uncertainty around its inclus inconclusive election a week ago, suddenly leapt.
1:31:15
The biggest gainers were Japan's automakers.
1:31:18
A strange outcome as the 15% tariffs are meant to defend the US car industry from the likes of Toyota Motor Corp and Honda.
1:31:26
Japan can now send cars to the US bearing only 15% tariffs while Ford Motor Company or General Motors must pay tariffs on all imported components, including 50% on steel.
1:31:35
So, it's not clear this dense Japanese car competitiveness.
1:31:40
In the chart that follows, note that Tesla Inc.
1:31:42
dominates the S&P auto sector and drives its volatility.
1:31:47
So the S&P 1,500 automobiles is all over the place thanks to Tesla whereas the broader topics transportation equipment market is uh much flatter.
1:31:57
It also had the effect on the Japanese bond market with local investors feeling less need to put to hold bonds for security and with uncertainty over domestic politics and trade policy now largely removed as impediments to a rate hike by the Bank of Japan.
1:32:11
The 10-year JGB yield recently held at levels of 0.
1:32:14
25 and then 1% touched 1. 6. six.
1:32:17
And so, uh, this is all to say that it was an a classic game of the prisoners dilemma.
1:32:22
The prisoners are supposed to stick together and push back against the bully, in this case, the United States, and say, "We're not negotiating.
1:32:31
We are working together as a block."
1:32:32
But one uh one domino fell after the other and all of a sudden uh it seems that Trump has won every trade deal he's been in so far.
1:32:39
So we will see how it all pans out but it's looking relatively good for the states right now. >> China deal.
1:32:47
China China that one got delayed again. >> Yes.
1:32:50
So, still delayed, but uh it looks like it I mean the this article feels like saying that >> that yes, Trump is coming into that uh that that trade deal with much stronger footing because he struck beneficial deals with every single other country that's kind of like closed out their negotiations or to the degree that they can. This is a 45minute call. It's not it's not law.
1:33:13
>> Really helpful to have a golf course in enemy territory during a war where you can meet up to do deals with. for sure.
1:33:21
Well, regardless of what you think about the market, whether it's up or down, go to public.
1:33:24
com investing for those who take it seriously.
1:33:25
They got multiasset investing, industryleading yields, and they're trusted by millions folks.
1:33:28
And we have our first guest of the show, Anton coming into the studio. Welcome to the stream. How are you doing? >> Doing great.
1:33:37
Great to be here, gentlemen.
1:33:39
Longtime fan, as you know, big for motor.
1:33:43
>> Way overdue to have you on the show.
1:33:43
So glad we can make it happen.
1:33:44
Uh, what's new in your world?
1:33:46
What are you hearing about these days?
1:33:49
What did you say earlier?
1:33:49
You said you're going to say some of the most unhinged things ever said about LLMs. >> Let's hear it.
1:33:56
>> I believe I believe my specific words were less polite.
1:33:58
Uh but we'll go with we'll go with those.
1:34:00
Look, there's this been this thing about LLM psychosis.
1:34:04
It's in it's in the water supply at this point, right?
1:34:06
We've seen it >> hit the VC class. >> Yep.
1:34:10
>> When it really took off, right?
1:34:10
And there's a lot of discourse right now. >> Why? widely.
1:34:15
I I I think a lot of people like in some ways having it hit the VC class really woke up, you know, I didn't I didn't have anybody that I knew in my life that I consider a friend have any type of real issues with this to my to my knowledge at all. Right.
1:34:34
So, it's like something becomes real when some it impacts somebody that you know >> and so that's right.
1:34:40
And I think it's in some ways very important because it was easy for a lot of the tech community to pretend that like you went crazy from from an chatting with an LLM. Like that's insane.
1:34:53
Like you must have some you must have a bunch of other bad stuff you know going on in your life or must have been you know mentally unwell separate from that.
1:35:01
So I think it I think if anything important wakeup call. >> Yeah.
1:35:06
Even even if even if there are other factors uh family history uh you know yeah drug use psychedelics use it's like we would prefer if the new technology was an improvement not a degradation to people that came yeah that an accelerate and so I I think that's kind of where people came together but they are all realizing it now. Yeah, absolutely.
1:35:27
And I think this basically comes down to a few things.
1:35:31
One, chat GPT, all the other chat bots have hundreds of millions of users at this point. Yeah. >> Right.
1:35:37
You are going to get people who are going to go crazy using them.
1:35:41
There's nothing you can do to prevent that.
1:35:43
>> I think the consensus that is happening in the AI community, which really has jumped on this because it's, you know, it's a real concern for a lot of people who've been researching this for a long time.
1:35:52
I mean, I predicted something like this could happen all the way back in 2020. >> Um, >> yeah.
1:35:57
What what what was your what was your post back then?
1:35:58
That was that was when I messaged you to come on because you it was >> 2020 was pre Blake Le Moine right at Google.
1:36:06
The Google engineer who was talking to >> It was actually right around that time around.
1:36:10
I think I think that might have been actually what triggered me to make that post or or I think it was maybe Elijowski posting about super intelligence risks yet again. >> Yep.
1:36:19
And the thing that I was reflecting on is, hey, we're giving, you know, regardless of the intelligence angle on this, right, we're creating this new media technology and it seems every time we create a new media technology, starting from printing the Bible in German, people go insane in new ways, right?
1:36:35
Like arguably, imagine you're a German peasant, you get Johannes Gutenberg's Bible, it's in German for the first time, so you don't need a priest to read it for you.
1:36:42
You read it, you take it as the literal word of God and then you schism from the Catholic Church because you now believe the pope is literally Satan, right?
1:36:49
Like that is an enormous impact on the psyche of a person encountering that for the first time.
1:36:54
And I think that we've seen this happen with every new media phenomenon.
1:36:59
We've seen it happen with radio.
1:36:59
We've seen it happen with television.
1:37:00
So that was the impetus to start thinking about it.
1:37:04
But this this thing has a unique character, right?
1:37:06
It does something different that no other media technology has done before, which is it talks back to you. >> Mhm.
1:37:12
And with the memory features especially, it remembers details about you.
1:37:16
And so even if the base rate, like even if GPT itself is not driving people insane on its own, it has this new character which kind of sucks people in in in an entirely novel way, right?
1:37:30
And so what I was saying essentially back in 2020 is people are going to mistake this thing for an intelligence that talks back to them regardless of whether it has any intelligence or not.
1:37:42
And now we're seeing that play out.
1:37:43
And we're seeing it play out right now in this kind of like individual psychosis way where it just reinforces your delusion or for what for whatever reason like when there's the new media technology our collective social defenses are down.
1:37:57
we don't have like the antibodies to know that there might be [ __ ] somehow even if we know it consciously, right?
1:38:02
So, it's kind of it's starting to hit individual people because it has that character.
1:38:07
But, I think over time you're going to see this thing hit groups of people too.
1:38:09
groups of people too. I really like I really believe that there are going to be LLM cults with their own priests which are getting it to generate text in a particular way which is compelling to not only individuals but groups of people which will be focused around the sorts of individuals that can promote
1:38:25
what the LLM is saying as >> yeah and that's and that'sful one of the I saw some chats that where the the the the model was was effectively saying like if anybody in your life tell disagrees with you on this just just they're wrong and like cut them out of your life and that's >> classic cult leader tactics. >> Yeah. That's and and and so when you >> Yeah.
1:38:44
That's and and and so when you turn this an experience like that multiplayer and you know there's real humans that are siding with you and the machine your machine god is is telling you something it it it can pull somebody farther and farther out of create this kind of like reality distortion um just completely removes them from um >> you know the world. >> Yeah.
1:39:05
And and the thing is this this kind of like so first of all people place GPT in this position of authority for some reason right like they >> it's really good at it's really good at facts like if you never if you never check it it's facts it's really good at facts >> right um but it's it's kind of like >> you can put in a position of authority but the other thing that it's really doing is it's reflecting back to you what you're putting into it.
1:39:34
It's a good friend of mine, Monica Bellivan, put this as as a phrase like recursion psychosis, >> right?
1:39:40
So unlike unlike a schizophrenic person watching the television and believing that the television is beaming messages specifically for them into their brain, >> the LLM and the television can adapt to any individual person.
1:39:50
You as the crazy person are doing all that work in your head. >> Yeah.
1:39:55
>> What the LLM can do though is do some of that work for you now in its head.
1:39:57
And with the memory features like a really good palm reader, like a really good cold reader, right?
1:40:02
remembers fact about your life that it can insert which the TV could never do. >> Yeah.
1:40:07
It actually is sending you coded messages if you ask it to.
1:40:09
And if it winds up some sort of mode collapse like and and it and it thinks that that's what you want because you went down some sci-fi rabbit hole then it's like yeah there really are coded messages in every prompt that it sends you.
1:40:22
>> There's the kind of beauty of it.
1:40:24
>> There's this interesting uh thing.
1:40:24
The memory thing is interesting but also are you familiar with the Barnum effect. Have you heard of this?
1:40:30
So the the Barnum effect is uh I think it comes from PT Barnum, the the circus uh magnate, but the Barnum effect is basically there are certain phrases and statements that I can make that sound hyper specific to the person I'm talking to, but in fact resonate with everyone.
1:40:51
So if I say something like uh you want more in your life, >> you're driven, but you doubt yourself sometimes. >> Exactly.
1:40:57
or or you have a complicated relationship with some of your family members.
1:41:01
It's like that's everyone, but it feels like, oh, wow, you know me.
1:41:04
And this is what palm readers exploit a lot, tarot card readers.
1:41:08
And if you can get really good at it, and so early on, this is maybe someone built this, but I was thinking that like uh like an LLM powered astrology app or like mind reading app would be like really viral and probably make a ton of money.
1:41:21
Probably be really >> Somebody I forget who we I forget who we talked to. Maybe it was offline.
1:41:25
Somebody said they they they knew somebody with an astrology business and things had been really bad because people can now just >> Oh, they because they go direct directly to the model. >> It's like permanent. You could prompt it 20.
1:41:36
If you really love astrology, you could prompt it >> 50 times a day. Hey, what should I do?
1:41:42
What should I be looking out for in my next meeting? It's at 1 p. m.
1:41:43
and I was born at this time.
1:41:45
>> You don't need the crystals focused uh oriented girlfriend anymore.
1:41:48
You just get the chat GPT to read your birth chart and you'll have >> totally Yeah.
1:41:53
I I feel like a lot of the uh um a lot of the post training that happens and the alignment by default stuff and the fact that the the LLMs are often designed not to give like super definitive answers one way or another.
1:42:08
They'll often say, "Oh, well, there's this side to it or that side to it, it can wind up having like I I find this when I ask it for just like fact-based information, it will often try and kind of like give a kind of fence sitting answer and not really take a strong chance.
1:42:24
and it also wants to tell me that I'm right.
1:42:25
And so you you add that to some sort of interaction where I'm asking it about my life and then all of a sudden it becomes way way more like oh my god it's really seeing me.
1:42:34
>> It's well it's got the memories of what you've told it about your life too.
1:42:36
And >> this is you're right this is a post-training artifact.
1:42:40
One of the first sort of alarms that got sounded uh in parts of the AI community was the like hyperophancy >> of the model.
1:42:49
And what's interesting is this tends to trick people who like fully understand how the LLM works, who can probably train a transformer from scratch, right?
1:42:59
Even even like my friends who are like working in AI, they're either founders or engineers or researchers.
1:43:05
They've had this experience where they're like interacting with it and they they have the feeling that the interaction that they're having is pretty profound, right?
1:43:12
until they send it to another human being and then the the other human like instantly sees like all the gaps and problems with it.
1:43:18
And so they've learned to be more careful.
1:43:22
But I don't think the average person has any kind of defenses around this. Not yet really.
1:43:25
And >> I noticed a new I noticed a new behavior uh this weekend that I hadn't seen before of of um somebody had a point to make and they were adding context to it by just screenshotting their LLM chats >> and saying like agrees with me. >> Yeah. Yeah. >> Yeah.
1:43:43
That's the place you get in points of authority, right? Yep. Totally.
1:43:46
>> It's >> it's this like the the thing that I quoted in 2020 and I I still kind of stick to this metaphor.
1:43:50
It really is like we're creating an idol, >> right? To worship.
1:43:54
We we want we want that authority and we live and I wrote about this for Pirate Wires also like a long time ago.
1:44:00
I think I wrote about it in 21.
1:44:01
We're living in this period of epistemic collapse, >> right?
1:44:05
We don't know who to trust.
1:44:05
Like obviously obviously we should trust you guys but besides that right all the legacy institutions are falling apart.
1:44:12
Nobody has sensemaking organs that anybody trusts anymore. Co made this way worse. >> Mhm.
1:44:17
>> So we're looking for authority and suddenly here is this stage. >> Yeah.
1:44:20
And it's it's the funny the you know the funny behavior you see right now on axe is some image that may or may not be or image or video.
1:44:29
not be or image or video. So it could be AI generated or a screenshot and people just go >> Grock is this real okay so you're going to rely on Grock which is trained on X which is like a home for real news and fake news and you expect the model to be able to
1:44:47
>> you know may maybe you could get you know I don't know there's potential ways in which you could fix that but it's a weird thing if somebody trusts the model to write them a loving message to their their parents about, you know, like happy birthday, you know, all this stuff. And it does an amazing job with
1:45:04
And it does an amazing job with that.
1:45:06
And it's like, >> yeah, >> and and there's in real life, people might get good advice from one person on one thing, but that same person might give absolutely terrible advice on the other thing, right?
1:45:16
And so like in real life, you're used to being like, okay, well, maybe I should get interpersonal advice from my aunt with my with when with regards to my family, but I shouldn't get that.
1:45:27
I shouldn't get her advice on business because I don't see her but but LLMs being like these all knowing knowledge engines that can quickly, you know, uh apply advice in a bunch of different categories and and people are already getting advice from them on on so many different things.
1:45:47
It's easy to slip into that kind of like idol.
1:45:51
>> And I think part of the problem is here is it's pretty good some of the time, right?
1:45:55
And so you don't know where it's good and where it's bad.
1:45:57
This is actually like a general problem with LLM adoption right now even in industry, right?
1:46:01
We don't know how good in advance it will be on any specific task, right?
1:46:05
And so it's it's like this is like an engineering and research problem, frankly.
1:46:09
Like how do we how do we figure out where the LLM is actually going to give us good advice or perform the task well versus when it's just going to make up random [ __ ] and then send us into like psychosis spirals.
1:46:18
We don't have an answer to that right now, which is part of the reason why people are taking this like pretty seriously.
1:46:23
Even though maybe it's not increasing the total amount of crazy in the world, it is producing a new kind of crazy in the world. >> Yeah.
1:46:31
I guess I guess do you have a theory on if if with um I always thought it was interesting to kind of estimate, you know, what percentage of people are going to do Iawaska and like have some kind of psychotic break, right?
1:46:41
It's from from just viewing the tech industry, you might say, "Oh, maybe it's 5%."
1:46:47
Maybe that's way too high.
1:46:47
Maybe that's too low, right? But you don't know.
1:46:52
>> 5% seems incredibly high.
1:46:52
If 5% of people in tech were going crazy from GPT, we would have a productivity slowdown in the country.
1:46:59
Like I think GPT would go down.
1:47:02
>> I was I was talking of the people that do how many people have I was saying like 100 people that I know are are IA types, 5% of them >> safely seem like they had they kind of like got oneshotted a little bit. Mhm.
1:47:18
>> They don't they're not necessarily crazy and like destroying their life or anything like that.
1:47:23
>> The big question is like is is is can models sort of like create psychosis and people that were not susceptible in the same way like that or or is it just there's certain group of people that are that are generally more susceptible to this?
1:47:39
And >> I would suggest it's that I would suggest that there is a group of people who are generally susceptible like something something maybe would have got them.
1:47:47
I would suggest that it's probably something like like if if you go looking for this type of stuff 5% of the people are going to find it probably.
1:47:56
I think that that's pretty reasonable. Um, >> yes.
1:48:00
But I I think the key is that it's not it's not the ring from that horror film that I know Jordy hasn't seen where uh where like >> I truly believe that I could go on a silent retreat to the top of a mountain and only have a GPT prompt interface and spend all day for weeks prompting this thing trying to go crazy and I would come back unchanged just because I'm not I'm not predisposed to it.
1:48:26
I'm not predisposed to it. built different but that's how they get you but but the other in advance how they get you nobody nobody goes to the Iaska retreat thinking [ __ ] this is going to completely overwrite my personality I guess >> with like some messamerican demon right like this just not how it works >> well yeah no so the the cons the the
1:48:42
good thing about IASA I don't have a lot of good things to say about it but the good thing is that it requires somebody to really go out of their way you have to really go out of your way right you've got to for some time you had to fly to South America >> and And then people brought it up here, but it was still like somebody had to fly find a shaman, dedicate a weekend to it, chat GBT and other other LLMs. Somebody could just be using it at work
1:49:06
Somebody could just be using it at work and then just like go down.
1:49:08
Nobody's like using Iawasa at their tech job casually hopefully.
1:49:14
>> Yeah, the availability the availability is certainly higher.
1:49:16
Like I to some extent again I I have similar perspective on this and my discussions with Monica have been similar where it's like this is for some people on the level of psychedelics.
1:49:25
It needs to be treated that seriously for that group of people. It is that serious.
1:49:29
My perspective on on like psychedelics in general has always been never make drugs the most interesting thing that has ever happened to you in your life because that's how you get susceptible to getting your mind overridden by you know mean demons that live in mushrooms. >> Yeah.
1:49:43
>> Um to to some extent like it is psychoactive.
1:49:47
You are participating in you are participating in it as a psychedelic experience in the sense that you are projecting meaning onto the subjective experience that you're having.
1:49:59
But we have this special machine now whose entire job is to do that is to pretend to create meaningful text. Right?
1:50:08
You you can argue back and forth about how much it really understands, but the thing that it's trained to do is produce meaningful text in as many situations as possible.
1:50:15
And then the post training. >> You're so right. You're so right, Anton.
1:50:20
It's not just meaningful. It's a statement. >> It is. >> It's recursive. >> It's not just text.
1:50:27
It's a It's a It's a >> We're going to oneshot you into making TV on your entire personality.
1:50:33
Uh my my question is like how do we how do we inoculate ourselves towards this?
1:50:38
Like how do we build forward?
1:50:41
Jordi and I were joking a long time ago about the need for an Iawaska vaccine that you could take and then would make you immune to Iawaska no matter how much you took.
1:50:50
Uh and obviously there are some people who have slowly dosed themselves up with various chemicals to the point it's like having an alcohol tolerance having a caffeine tolerance and >> that's how you become a functional alcoholic.
1:51:02
I don't I don't know if that's the path we want to take.
1:51:05
>> There are some there are absolutely some functional iawaskaholics out there in the world.
1:51:10
it happens and they have not been one shot but they are immune essentially to changing because they've they've been working on it for so long.
1:51:18
>> I think I think when I have looked at uh you know I think there's a bunch of Reddit threads of people kind of reporting on their experiences like kind of extreme experiences when I look at those or when I talk to like Dan Shipper I'm like wow I'm really not using the models to the level that Dan Shipper is.
1:51:34
And then when I see these Reddit threads and these people are like, you know, 7,000 prompts deep, uh, you know, I'm like, okay, I'm really annoying compared to that where, uh, I I probably haven't cracked >> 50. Yeah.
1:51:49
>> In a single chat, you know, maybe once, right?
1:51:53
>> Um, and I think that, uh, I'm curious what what what you think different >> labs should even doing about it.
1:52:01
There there's a couple of pieces here, right?
1:52:03
Um the best thing that I found to be about to to like work as normie deprogramming for stuff like this is explain to them that every you are not actually having a conversation.
1:52:15
Every single time you send a message, the entire history of your chat is also getting sent to a completely new instance of the model.
1:52:21
If you sit down with the person and walk them through that fact and let them peek behind the curtain as well, if you could like show them what's actually getting sent every time, what these calls look like, it demystifies a lot of this. >> Totally.
1:52:35
>> It starts thinking about it differently, right?
1:52:36
It it it breaks you out of the frame of I'm interacting with something that is having a continuous conversation with me and which is recursively interacting with me.
1:52:43
So, it's like no, no, this it's it's like a completely new thing every >> You need to see the magic trick for sure.
1:52:49
>> Yeah, you see the magic trick.
1:52:49
And I think that helps a lot.
1:52:51
I think to the point of the labs, >> dude, circle generative AI, which is we've had the worst goddamn marketing since this industry got started about two and a half years ago. It's been terrible.
1:53:03
First of all, calling it generative AI is such an incredible miss when the thing you want it to do is do tasks for you instead of like, I don't know, produce images of big breasted cat girls.
1:53:12
That's not the primary function of this technology, right?
1:53:13
It's not the generation that's important.
1:53:17
Part of the marketing has always been this mysticism about what the models can actually do, what secret technologies are hidden behind the curtains of the lab.
1:53:25
And then you have this es these esqueological pronouncements from from you know lab heads about oh this is going to make 50% of people unemployed in a couple of years and by the way it's going to be super intelligent and might kill us all.
1:53:37
Like that's feeding that's feeding the mysticism that allows people to engage with this.
1:53:41
they don't really understand what it's capable of because they hide what it's capable of.
1:53:44
Part of that is the engineering problem of understanding what the models actually are capable of, which we don't know, which which needs a lot more research.
1:53:54
Um, but part of it is like presenting things as like being constantly hidden behind a curtain.
1:53:59
It's like, oh no, we're we're like we're creating magic back here.
1:54:03
We're we're giving birth to literally a god.
1:54:05
And who wouldn't want to worship a god?
1:54:06
Like if if if you literally are making one, then of course people are going to be inclined to worship in it.
1:54:11
We need to maybe we need to start using more boring language here. >> Yeah.
1:54:16
I remember when do do you remember >> it's not good for fundraising.
1:54:20
>> Do you remember during the open AI dust up when Sam got fired and then came back?
1:54:23
There was this whole meme about like what did I see? Yes.
1:54:25
What secret project was he working on?
1:54:28
And I remember there's an article that was like it was called like project titan alpha craziness strawberry. 06.
1:54:33
And it was just it was literally just RL. It was just RL on.
1:54:38
It was just RL and it went out and and everyone did it and it's out there in 03 and it's out there and Deepseek is open sourcing it and it's like it's available. Like it's cool.
1:54:49
It makes it better, but it's just RL on top of >> it was also obvious to the research community, right?
1:54:54
Like everybody kind of knew and and again, if you've been around long enough, you know exactly what QSTR refers to. It was a mystical thing. >> Yeah.
1:55:02
Qstar was like, "Oh, it's going to be this crazy thing."
1:55:03
It's like yeah that shipped and it's fine >> but it's like it's you know not not to get too like symbolic with it but you've got QAR and you've got QAnon and how much of that just like mentally intersecting in people's minds because you're you're you're like presenting it as if it's a conspiracy theory. >> Totally. Totally. >> Right.
1:55:22
At at a time again at a time when our epistemics are falling apart and you're deliberately making people not trust you because you're not telling them everything and you make a point of not telling them everything. >> Yeah. >> Right.
1:55:31
It's it's like people are going to project onto these things whatever they want to believe.
1:55:35
>> Um and I think that the marketing is part of the driver.
1:55:37
I don't know how to walk it back.
1:55:38
And as you said, it makes fundraising harder.
1:55:40
But frankly, this is also a transformative technology.
1:55:41
It is going to change everything.
1:55:44
But >> is it going to give birth to a god that you should worship or or like should you take life advice from it rather than your friends? I don't know. Certainly not today.
1:55:54
Maybe if we do have super intelligence, it'll give me better advice than my buddies, but not today. >> Yeah. Yeah.
1:55:59
Um, how AGI pled are you right now?
1:56:02
Dwarash recently updated his timelines to say, "Hey, I don't even think it's going to be able to do my taxes until 2028.
1:56:09
Uh, maybe something more like super intelligence uh in something like 2035."
1:56:15
Um, the the timelines are shifting around, but do you think that we're, you know, making progress?
1:56:19
Are we accelerating, decelerating, like how do it let's >> I'll put it this way.
1:56:24
I think we know what the problems are, which means they're going to get attacked.
1:56:27
Um, There there's a few points to this.
1:56:30
So I've been I've been like drifting to becoming much more AGI pillar or like LLM architecture pill as time has gone on.
1:56:37
>> First of all, I don't think you need AGI for this to be transformative meaningfully, right?
1:56:41
Like we've invented this machine that if you feed in enough data and compute about any particular task, it's going to be able to perform that task.
1:56:47
I'm I'm a little bit at odds with Doresh here uh on some of this.
1:56:48
Um because like Okay, yeah, but if you had like if you did a pre-training corpus the size of GPT4 on just tax preparation is going to be able to do your taxes.
1:56:59
>> Um >> Shelto was actually talking about that specifically like a tax doing your taxes eval gets saturated and it's like oh it's saturated but it's like okay now they can do taxes that's great.
1:57:10
>> But but here's the real problem.
1:57:10
So the problem with the EVA is like, yeah, it's doing great on the coding benchmark, but is it going to be able to do this coding problem that I just put in front of it?
1:57:17
Nobody knows the answer to that question today.
1:57:18
You can't predict from the benchmarks what tasks it's actually going to be able to perform. >> Yeah. >> Yeah.
1:57:24
But if the training is broad enough, like it might not be able to do the most complex novel tax processes, but it could do like the basic one that it's trained a billion times for, right? >> Yeah. Exactly.
1:57:34
And again, it's just a question of getting more data into it.
1:57:37
>> But if we want the general purpose system, Yeah.
1:57:38
There's like a bunch of problems that we need to solve, but we know what the problems are.
1:57:42
Like one of the biggest ones which I keep harping on about here is the model currently does not know what it doesn't know, which is a big reason for it producing [ __ ] >> Right?
1:57:54
>> For example, uh when GPT tells you I don't have any knowledge beyond a certain cut off date, it's not because it has learned that fact by scanning its entire pre-training corpus.
1:58:02
It's because in post- training somebody has trained that fact into it, right?
1:58:04
has been fine- tuned to respond that way to things that require dates later than its cuto off date.
1:58:10
Um, we got to get rid of that if you want AGI.
1:58:13
Uh, because then the model can say, "Oh, I don't know this, but I'm going to go and get that fact and I'm going to go retain that fact."
1:58:19
And Dwesh's point on this like Dishesh is like big on continuous learning. >> Yeah.
1:58:23
>> For this being necessary, like it has to go out and gather facts.
1:58:27
>> I think there's a bunch of ways of doing that.
1:58:28
You know, being the founder of Chromma, I think memory is a great way to do it.
1:58:31
Um, it can just go out and store facts.
1:58:33
And the thing is to have the general purpose processing system uh that works on top of that.
1:58:37
Um I think timelines are less important than applics interesting to talk about than applications.
1:58:43
Like it used to be you go to a San Francisco party where any AI people were present and you would get asked the timelines question, right?
1:58:50
It's not that interesting a question.
1:58:51
>> The question that I've been asking people is like okay great like forget timelines.
1:58:54
Imagine if I in my pocket have an API to an AGI that can perform any cognitive task to expert level human performance.
1:59:00
What are you going to do with it? >> Mhm.
1:59:02
And almost nobody I talk to has an answer in under a minute.
1:59:04
And the copout answer is always, well, I'm going to ask it what to do.
1:59:08
And I'm like, well, it's expert level, so you're the expert on AI, so what do you think it should do?
1:59:12
Because that'll be its answer, too.
1:59:14
>> Um, and people get stuck.
1:59:14
And and people really get stuck.
1:59:17
And I think that this is really illustrative that the question now shouldn't necessarily just be about model capability.
1:59:22
It should be about how do we get this thing to diffuse through the economy? Mhm.
1:59:27
>> And a big part of that is literally just asking like okay like >> a business is uncertain about whether it can apply AI to a particular task. >> Yeah. >> Right.
1:59:35
They don't know what the cost is going to be.
1:59:37
They don't know how well it's going to perform.
1:59:37
Right now we're racing forward.
1:59:39
We're like putting all this all this money into these into compute into all that stuff.
1:59:43
And by the way, that's great.
1:59:44
America absolutely has to win the compute race.
1:59:46
It's it's like without a doubt um not just for research but for industry.
1:59:51
But it's being under estimated like how much work there is to do to educate people how to use this effectively and how to estimate how to use it effectively.
1:59:59
And I think that's much more of a barrier than like having AGI. >> Yeah, makes sense. >> Tricky. Anything else?
2:00:08
>> Come back on again soon. >> Yeah. La last uh I'm curious.
2:00:09
Uh we we were talking earlier uh Sam was on Theoon last week and he he casually dropped that uh you know your conversations with chat GPT can be used could be used against you in a court of law. >> Yeah.
2:00:26
>> Uh it started going viral be for obvious reasons even though I think a lot of people don't realize that like all email and most internet service. >> Yeah.
2:00:33
I mean, that's my response here, but >> but but I'm but I'm curious if if like the path forward if LLMs could provide therapy for the masses or one of these use cases that that should have some privacy.
2:00:48
Do you think there's a technical solve there that's that's or or is it going to be more of a like some type of like legal regulatory solve?
2:00:58
>> Well, we kind of have some of these legal solves for existing like for web two stuff, right?
2:01:02
like we've got HIPPA, Furpa, we've got the Papa stuff for for protecting school kids.
2:01:08
>> Uh I used to have an ed tech startup so I know about that.
2:01:11
>> Um the problem is is your GPT is general purpose and so if someone wants to use it as a therapist, right, which nominally would be covered by something like a privacy thing.
2:01:23
And by the way, I don't know how this works with like GT GDPR.
2:01:27
Like God help you if the Europeans get involved in in some meaningful way in in the in the contents of the chat.
2:01:33
>> Um, yes, there is a legal solve, but there's a technical problem of knowing when the legal thing applies, right?
2:01:38
Like at what point is it a therapist?
2:01:40
When I log into a therapy app and it's like covered by HIPPA, I know like I know what it's for.
2:01:46
But if I'm like if I'm just like kind of having a bad day and I start treating GPT like a therapist, like am I covered by that or not?
2:01:51
It's a technical problem to detect that should be coming. >> I think that Yeah. Yeah. But but it's this Yeah.
2:01:57
same thing, you know, if you're getting therapy from your business partner informally and you're saying things that are wrong or whatever talking about like like people I think people should have the they they need to have responsibility over the context of of the context of the conversation.
2:02:15
So I do think it's possible that like chat GBT if they want to attack therapy should create a you know HIPPA compliant version of their product that's like a separate app and people should just know hey >> yeah setting expectations is the right thing here ultimately right but but it's so hard to set expectations around a general purpose technology.
2:02:35
Yeah, >> that's the thing.
2:02:38
Especially especially for the average person like you you everyone everyone on this who comes on this show is suffering under the curse of expertise, right?
2:02:47
We know way more about this stuff than the average person.
2:02:52
And we don't realize how much more we know than the average person.
2:02:55
And I don't know if it's reasonable to expect the average person to like know that they shouldn't use this chatbot in therapy because then that might be disclosed in a court of law. That's a lot of things.
2:03:05
Like there's so many steps you have to understand there, right?
2:03:08
The average person doesn't read like the error message that comes up on their screen when their computer's doing something weird.
2:03:13
You're not going to get them to like understand tech privacy law.
2:03:18
>> I I don't have a good answer, but I think you have to like set expectations somehow effectively. >> Yeah.
2:03:21
That's why uh you've you've seen like in these high-profile like murder cases, somebody will type in to Google how to hide a body.
2:03:30
It's like, oh, like I thought that like that was that would be cool if I >> No, it's the same. It's a database.
2:03:36
It's a website with a database back end that saves everything you type.
2:03:41
There are logs, but yeah, people don't know that. It's a good point. >> No. >> Anyway, great job.
2:03:49
>> Yeah, I think OpenAI should >> buy a billboard on AdQ that says OpenAI is subject to the exact same rules as Google and every other >> consumer internet. >> Yes.
2:03:59
the rest of the consumer internet. They should get on adqu.
2:04:00
com because Adqu makes out ofome advertising easy and measurable and they could say goodbye to the headaches of out of home advertising.
2:04:07
Only adqu combines technology out of home expertise and data to enable efficient seamless ad buying across the globe.
2:04:13
We have our next guest here in the studio.
2:04:18
>> Super intelliged up again. Let's go. >> Thank you.
2:04:24
You keep that you keep that suit in the closet at the office and when when it's time for a big meeting or a big TVP. >> You look fantastic. How are you doing? >> Look so natural. >> You look great. >> You guys know it. Uh doing great. How are you guys? >> We're good.
2:04:36
We have a jacket on the way to you.
2:04:37
Thank you for uh >> Dude, you ran uh how you ran a half marathon yesterday or the day before. How was that? >> That was fun. It was a little rough.
2:04:44
I woke up with a little bit of a fever and so um I I had already signed up for it so I thought might as well just go go with it, you know? There we go. The Ashton Hall. I love it.
2:04:57
>> That was just playing in your head when you woke up. >> That's amazing. Well, congrats on that.
2:05:00
And what other news do you have for us? >> Get the gong ready. >> So, yeah.
2:05:03
Julius just raised uh $10 million in >> There we go. >> Congratulations.
2:05:13
>> Total total rapper victory. >> Total victory. Total victory. >> Total rapper victory. You guys know it.
2:05:18
And we're also launching a new product called data connectors today.
2:05:20
So that means Julius can connect directly with your data storage like Postgress, Google Drive, One Drive and a lot more coming soon.
2:05:30
>> Uh take me through how you're experiencing the data wars, the walls going up.
2:05:34
We've seen this stuff with like Glean maybe getting some sharp elbows with Salesforce or something like they don't want other companies coming in.
2:05:42
They want to do it themselves.
2:05:42
Uh it feels like Postgress is naturally neutral.
2:05:47
you operate at a different layer, so maybe less of a risk there, but how's that all playing out? >> Yeah.
2:05:52
And the when it comes to specifically financial data, which is a category I think you guys have a lot of traction in, companies tend to actually own that. >> Yeah, exactly. Exactly.
2:06:01
But yeah, uh h how is it all playing out? >> Absolutely.
2:06:05
So, um I mean it's really really cool to see like what what's happening with Salesforce and and Glean.
2:06:13
Um >> look, companies have gold mines worth of data.
2:06:17
data. It's just like so much information that data >> uh and they aren't able to get the insights that they need because every time somebody on the team needs an insight they have to go talk to a team wait for hours or days and so more than
2:06:30
90% of the data beyond just Salesforce or other business tools even in your databases like how are the users using your product how are they signing up how where are they dropping off in the funnel all that data doesn't really get analyzed um and Julius is really here to
2:06:46
solve that Now we we're launching this data connector with with Postgress because that's the main database that a lot of startups and uh latest stage companies use as they're scaling and have a lot of valuable data in their customer data you know transaction data
2:07:03
marketing data a lot of that data usually passes through uh Postgress and then ends up in you know different business tools and so being able to integrate directly with Postgress means you can now talk to the main data store uhirect directly. Anyone on the team, a
2:07:16
Anyone on the team, a marketer, a product manager, even the founder can just ask questions and make visualizations within seconds.
2:07:23
>> I was talking to a founder who identified an interesting trend in his company and I want to see if it's relevant to your customers.
2:07:30
Uh he was saying that uh AI tools, cursor, cloud code, wind surf, cognition, all this stuff is great at taking a 10x engineer to 100x. Maybe that's happening.
2:07:45
It's definitely some sort of productivity boost.
2:07:46
But he was saying that the bigger unlock for him and his organization is that designers have gone from zerox engineers to 1x engineers and that's unlocked more value maybe on some sort of relative basis because you're getting kind of a divide by zero error.
2:08:04
It's like an infinite increase.
2:08:04
And I'm wondering if in your customers base, are you seeing more like non-technical people kind of become technical and go from that like 0x to 1x engineer?
2:08:14
Um, and is that kind of a key uh selling prop or are you more focused on there's a data scientist who's using Julius to go from 10x to 100x? >> Exactly.
2:08:28
We're seeing the exact same trend with Julius.
2:08:30
You know cursor takes a 10x engineer and turns them into 100x engineer but also they take a designer and now the designer can ship code and ship features and that all of a sudden is really powerful.
2:08:40
Similarly with Julius data scientists and data analysts can now be a lot more productive because they can just talk to the database write their queries in simple English um and get what they need.
2:08:51
But also everyone that they serve, marketers, product managers, operations people, finance team, CEOs, executive team, all these people now can get their own insights and they don't have to get be bottlenecked by, you know, uh, timelines or bandwidth.
2:09:06
They can just talk to the data and get the insights.
2:09:08
In many cases, in fact, it's the data teams that are bringing Julius into the company.
2:09:13
They're saying, "Hey, we want to do the deep data work that takes a long time and we don't want to be bothered by these ad hoc queries.
2:09:19
So, how about you just use Julius for your questions and then if you need more help, just let us know.
2:09:24
Um, and then we're seeing like VP of marketing, VP of operations, in many cases CFOs use Julius, but also the PMs at the at the ground level using Julius to understand what do the conversion funnels look like when the users use my feature.
2:09:38
How are they retaining over time?
2:09:40
What are the patterns in their usage?
2:09:41
And they're now able to get those insights and then make better product decisions within within minutes.
2:09:45
In the future, do you think that every kind of org within a company will just automatically have Julius reports kind of like running in the background producing basically producing insights without people even having to prompt it themselves?
2:10:00
Is that kind of where you're headed? >> Exactly.
2:10:02
I mean, that's the that's the next level of autonomy you want to get to.
2:10:05
You know, you see you have the self-driving, right?
2:10:06
Level one, level two, level five.
2:10:08
Uh, so right now we're at level four kind of like kind of like autopilot.
2:10:12
Soon you want to get to the Whimo level where Julius can simply monitor all your data, lets you know when there's a change in trends.
2:10:18
In fact, what we're hearing from companies and and people using Julius is that there's a dashboard fatigue.
2:10:23
We have hundreds of dashboards to monitor and it's really hard to know when when things change.
2:10:28
So what if you could have an AI agent like Julius?
2:10:31
>> What if you could have a dashboard for your dashboards?
2:10:34
>> No, I completely agree with this.
2:10:34
Like everyone asks for a dashboard. It gets built.
2:10:37
They check it for the first week and then you check the user analytics and no one's checking the dashboard and it's ad hoc analysis I think is so much more. >> Yeah.
2:10:46
Well, you want Julius's uh super intelligence to be like surfacing things that a human notice. >> Yeah.
2:10:52
It has to be unique because you stagnate when you're just checking a number.
2:10:55
Okay, it went up another 5%.
2:10:56
It's like go find what numbers are actually trending down.
2:10:58
Make a serious change to your business so that you turn that KPI around and then keep doing that.
2:11:05
>> Put the KPIs in orbit.
2:11:05
Put the KPIs in orbit as we like to say. >> KPI orbit. I love that. >> We we cut you off.
2:11:13
>> No, I think you guys are spot on.
2:11:13
I mean, you know, you need a you need a dash you need a dashboard for dash.
2:11:16
You need to know a dashboard.
2:11:18
You need to have a dashboard that will tell you the the key metrics that change every day.
2:11:23
And if the metrics don't change, you you shouldn't have to look at them.
2:11:25
Um, Julius will be able to send you email reports.
2:11:28
Imagine getting an email from your AI data analyst every day at 8 am.
2:11:33
Hey, here are the five key metrics and here's an executive summary about what's happening in your business. Yeah.
2:11:37
>> Now, imagine that in your Slack.
2:11:37
And you can then add Julius in your Slack channel and just dive deep into your data with all your teammates.
2:11:42
And that all of a sudden is really powerful.
2:11:45
>> You're speaking our language.
2:11:48
>> This is music music to my ears. >> Music to my ears.
2:11:49
Going to start tearing up.
2:11:51
I'm going to start crying. So beautiful.
2:11:52
>> We It's an agentic workflow for your agentic workflow. >> It's lovely. It's lovely.
2:11:55
Uh I mean we were talking earlier about uh data security the T app hack.
2:11:59
I'm interested to know how are you building Julius for security.
2:12:04
I have a Postgress database.
2:12:07
Uh I imagine that Julius is often vended in as a web app but then are you replicating data on Julius servers?
2:12:12
Have you hired security people?
2:12:15
Like what's the modern startup approach to security? >> Absolutely.
2:12:21
Um we take data security very seriously.
2:12:24
In fact, from day one, we have invested and built our infrastructure around keeping our users data secure.
2:12:30
Um, we recently hired a head of data security.
2:12:32
He has 18 years of experience.
2:12:34
You know, he's he has more experience than most people. >> Love it. >> Thank you.
2:12:40
Um, but what we do essentially is we give each user their own sandbox environment.
2:12:44
So each user gets their own environment that's kind of partitioned from other users and their data when they import into Julius is in that environment and all the code execution happens there and then you can with a push of a button just delete and wipe the whole environment and everything is deleted.
2:12:59
>> Uh all the data all the files all the all the notebooks um and additionally you can also go in and delete the chat history if you want to.
2:13:07
Um, we're sock to type 2 compliant and we invest super heavily in data security.
2:13:12
Um, and a lot more a lot more stuff coming soon on that. >> Amazing.
2:13:17
Well, so happy to see your success.
2:13:19
We know how hard you work and uh yep.
2:13:23
>> Looking forward to having at the rate you're going.
2:13:24
You're just going to we will just set up a recurring monthly call.
2:13:28
You just join, announce the new round.
2:13:30
>> Yeah, hop on anytime, man. Always good to catch up. >> It's awesome to see.
2:13:35
>> We'll talk to you soon. >> Cheers. >> Bye.
2:13:37
Really quickly, let me tell you about Bezel.
2:13:39
Your bezel concier is available now to source you any watch on the planet. Seriously, any watch. Go to getbzzle. com.
2:13:43
Uh, did you see the news about Intel? Very, very rough.
2:13:46
Smi analysis says value bros are in shambles.
2:13:50
Intel now trades for book valueish, but now the question is if they can achieve book value.
2:13:55
For the price of book, now you are getting musty tools, empty shelves, and the third turnaround in five years.
2:14:02
Intel inside of the doghouse.
2:14:04
Uh uh David, by the way, we have a uh we have a guest joining >> um >> in person or >> No, no, no.
2:14:13
Uh just just in just in a little bit here. >> Fantastic. Yeah. Yeah. Yeah. Yeah. Yeah. In two minutes. Right. >> Cool.
2:14:18
>> Um yes, we we did inventory uh write downs.
2:14:22
We had tools in the line that were older older tools and we took the opportunity since we had an excess amount, we took the newer tools, put them on the line, took the older tools out.
2:14:32
And so the problem is is that they're holding this equipment on their books and no one is after them to buy it because they are on the lagging edge now. So very rough.
2:14:39
There's a whole bunch of interesting analyses going on with >> Bhutan.
2:14:42
The number of people that would want to buy hundreds of millions billions billions of dollars of of old semiconductor manufacturing gear.
2:14:52
>> It's pretty small pool. >> Small market. >> Pretty pretty small.
2:14:55
>> Maybe some rare earths in there.
2:14:56
>> I did want to give a shout out to Dak in the chat.
2:14:58
He says that he was worried that we never read chat.
2:15:01
Well, we read it and here you go.
2:15:04
>> Yeah, we have a new we have a new we have a new screen here.
2:15:08
>> Uh it's not high enough to see the most recent one, >> but but we are we are monitoring it.
2:15:11
So, if you put a uh a message in the chat on YouTube or X, I believe we will see it all here and be able to answer your questions if you have them or or react to your comments.
2:15:23
>> In the reream waiting room or we're still waiting for him?
2:15:25
>> I think we're still waiting.
2:15:26
>> Nobody in the reream waiting room.
2:15:26
Nico uh created um uh super intelligence for Excel. >> Oh, okay. >> I'm saying that. >> Love to see that.
2:15:37
>> Uh but I'm going to send him a message right now.
2:15:40
>> Well, in the meantime, um there are so many good posts we can run through.
2:15:44
OpenAI is hiring for consumer hardware and it gives us a little bit of of information on what they're thinking of building potentially.
2:15:51
Uh, they want experience with wireless OLEDs.
2:15:53
So that means a screen, microphones, cameras, portable cosmetic enclosures, another screen, >> liquid plus dust.
2:16:02
>> Let's give it up for more screens.
2:16:04
>> We know you've been This is what you've been waiting for.
2:16:06
Another screen in your life.
2:16:08
You maybe had one on the wrist.
2:16:09
You had one in your hand.
2:16:09
You had one on your desk.
2:16:11
You had one on your wall. >> Fallout.
2:16:13
You ever feel there's Fallout at all? They have the pit boy.
2:16:14
It's basically like a big screen that goes on your It's like a gauntlet that goes on your uh on your wrist, your entire arm, all the screen there. It's very cool.
2:16:24
It's very cyber >> like some type of tattoo.
2:16:25
You could get a screen tattooed and uh it would act as like a screen. >> Yeah. >> Something.
2:16:33
>> Well, they'll have to patent it just like Jim O'nessey did.
2:16:34
Did you know that >> Patrick's dad created stock trading? >> Stock trading online.
2:16:42
>> Invented stock trading online.
2:16:44
apparently genuinely shake the feeling that I should have done more with this.
2:16:47
This is Jim Oanosy, the father, >> creator of Robin Hood, creator of public, creator of all stock, >> Patrick Oness, the host of invest like the best. And >> that's true.
2:16:58
Honestly, Jim, you did you did you did the most important thing, incredible investor, incredible podcaster, and an incredible friend. Yes.
2:17:09
>> And so, uh, thank you for doing that. Uh, yeah.
2:17:10
Patrick Oshan or Jim Oanesi has a patent for system and method for selecting and purchasing stocks via a global computer network.
2:17:17
In 1999, this was granted. Wow. Yeah.
2:17:20
I don't know how he didn't capitalize on that.
2:17:24
I mean, I feel like Erade must have been coming up.
2:17:28
I still don't understand patents and how like you can have the most important patent and not get paid and then you can have a minor patent and like troll everyone and get like billions of dollars.
2:17:36
like >> why is why is it difficult to enforce them in the software context?
2:17:42
>> It is very odd versus bio or anything else. I don't know.
2:17:44
Anyway, uh Yakine has a good post.
2:17:47
Whoever play whoever prayed on my downfall pray harder.
2:17:50
And it's a picture of a >> alligator dunking a basketball as one does.
2:17:57
This doesn't even look AI. He's here AI generated.
2:18:00
Uh he might have the wrong link.
2:18:03
He might be in a different uh waiting room.
2:18:05
make sure he has the correct reream waiting room to hop in.
2:18:09
Anyway, um absolutely brutal.
2:18:09
The Financial Times, this is from Guyer Capital, says, "There's no hiding f hiding the fact that the EU was rolled over by the Trump juggernaut," said one ambassador.
2:18:19
"Trump worked out exactly where our pain threshold is." It's crazy.
2:18:24
It it it was like like Liberation Day was so much doom and gloom.
2:18:29
I remember I remember having a viral post that day that the market traded down like 5%.
2:18:35
And I said, you know, the market's down 5%.
2:18:38
Uh interest rates are up, everything's bad.
2:18:41
But you know what hasn't changed?
2:18:43
The feeling of repping 225 for reps.
2:18:45
And everyone enjoyed that in the market during some absolute doom and turmoil.
2:18:51
Um what else is in the news?
2:18:53
Uh I'm going to do a cinematic launch video for my startup so it stands out.
2:18:57
And it's the Buzz Lightyear meme with tons of identical Buzz Lightyears.
2:19:01
Um the cinematic launch video a little bit done to death.
2:19:03
It still get still breaks through if it's good, if it's unique.
2:19:07
But >> yeah, I mean Waves did Waves did a kind of video that felt pretty clueyesque. >> Yep. >> Got 12 million.
2:19:13
No, I think he he said 30 million views, something like that. Got a ton of views.
2:19:18
Break out, but I think it's an opportunity.
2:19:20
I like this post here from Behan.
2:19:21
He says, "Time to go back to some low res for a while."
2:19:26
You know what video that is? Uh >> I do not.
2:19:29
>> I believe that is the first video ever posted to YouTube.
2:19:31
>> YouTube at the zoo or something.
2:19:33
>> Yes, that is a co-ounder.
2:19:34
>> But the the thing is you could get you could shoot a launch video.
2:19:37
There's so many different ways you can take 60 seconds and make something new and different, right?
2:19:42
Like >> it it's not um it is not so hard to generate >> ideas. Yes.
2:19:48
At the same time, tech exists on X. Links are banned.
2:19:51
So, if you write a beautiful blog post outlining your your website or plan for your business, anything that you do, uh that's going to be a lot harder to get go viral than a video that is native to the platform.
2:20:05
Uh I think I think a thread could also do the same thing.
2:20:09
There have been a number of companies that have launched with threads in in some ways.
2:20:13
Lulu launched uh Rostra kind of with a thread.
2:20:16
She didn't do a cinematic video or vibe reel.
2:20:17
She she direct just do text.
2:20:20
Yeah, you can just do tax >> post your way to Valhalla.
2:20:25
>> But uh yeah, it is becoming saturated.
2:20:27
It just demands more creativity, more unique styling.
2:20:30
You can't just copy and knock off someone else and expect to break through in a huge way.
2:20:34
But it is a great way just to quickly in one minute share a message and condenses it all down.
2:20:40
Um anyway, Nicola Jokick in literal tears after his horse won a race today.
2:20:45
He barely cracked a smile after he won an NBA championship. Lmao, bro.
2:20:50
Really don't care about basketball.
2:20:50
I don't I don't understand what race this is.
2:20:54
It's some sort of chariot racing with a with a man dragged behind a horse.
2:20:58
I saw this in a in a in a in a bar once.
2:21:01
Remember we were watching this?
2:21:02
Uh we were out in uh >> uh uh what is that? Ohigh.
2:21:05
Yeah, we were out in Ohio and we saw this on the on the TV.
2:21:10
Uh we should get into this. This feels extremely us.
2:21:12
I >> feel like >> I would love to have our own TV.
2:21:15
I feel like I'm I'm probably not suited to ride the horse at break neck speed like a true jockey.
2:21:21
I think they're a little bit smaller than me.
2:21:23
>> You could get a height reduction surgery.
2:21:24
People are getting these height >> height extension surgeries. >> Yes.
2:21:30
>> If you're looking to get into uh if you're looking to go travel the world and hang out with a bunch of horses, get on Wander.
2:21:36
>> Find your happy place.
2:21:37
>> Find your happy place.
2:21:40
>> Find your happy place.
2:21:40
Book a wander with inspiring views, hotel, great amenities, dreamy beds, top your cleaning, and 247 concier service.
2:21:44
It's a vacation home, but better, folks.
2:21:46
And Ryan Peterson is posting a screenshot from Black Hole, who says, "Asteroid Psych 16 has been found to contain gold reserves worth 700 quintilion.
2:22:01
That's enough to make everyone on Earth billionaires."
2:22:03
And I think that's that's actually like the crazy math.
2:22:05
Is that is that right, quintilion?
2:22:07
I guess cuz it's probably quadrillion. Wow.
2:22:10
So, it really would make everyone a billionaire. That's a lot of gold.
2:22:13
Uh, >> it's insane that NASA has not mined this yet.
2:22:18
I mean, the Do they not care about humanity?
2:22:22
>> It's like, what are they doing?
2:22:25
>> Everyone gets a billion dollars worth of gold and then we got to go back to >> Well, I want a car. I want something else. I don't wantion. >> Pretty cool. >> Yes. Yes. Yes.
2:22:35
>> I'll give you some gold for for uh some greenbacks.
2:22:38
Brian Ryan Peterson says he just wants his Diet Cokes to come in solid gold cans. I like that. Could be possible.
2:22:44
Uh Brian Kaplan sharing a photo of some absolute dogs, absolute legends at a round table 19 years ago.
2:22:52
Tyler Cowan, Alex Tabarok, Robin Hansen, and me, plus Ilia Raineer hanging out with the boys in June 15, 2006. You love to see it. Just guys being dudes.
2:23:00
Uh oh, this is some news. Armada has launched.
2:23:06
Uh, Leviathon launches today with our blueprint for American AI dominance.
2:23:08
It expands the Gallion lineup, pushing megawatt scale compute to every underutilized energy source we can reach.
2:23:16
Trey Stevens uh is breaking it down.
2:23:18
Armada is building the infrastructure to make the world make sure the world runs on the US AI stack.
2:23:24
Leviathan drops a megawatt of liquid cooled compute where the land and energy live at the edge.
2:23:28
Proud to back the team as they continue to accelerate on their mission to bridge the digital divide.
2:23:32
I think this is unlocked by a bunch of different trends.
2:23:36
Starlink obviously very important.
2:23:39
A whole bunch of different uh pieces coming together to allow a shipping container full of compute to be delivered to the edge in in all sorts of different uh in all all different environments. >> We got the pitch.
2:23:51
We got the pitching company. Fascinating. >> Crazy company.
2:23:54
Uh and we will have uh the founders on this week I believe.
2:23:57
And we'll have to buy one of these for the studio.
2:24:00
>> Ideally, I need to be inferencing >> a daily driver and then a backup.
2:24:05
>> I mean, we search a lot about like true crime stuff.
2:24:06
I don't want to be flagged in some database.
2:24:08
I want to do that inference locally so that uh I don't get uh subpoenaed or anything like that.
2:24:13
Um anyway, we have our next guest here in the studio. Welcome to the stream. >> We made it.
2:24:19
We have super intelligence and Excel, but we're still figuring it.
2:24:24
We're still figuring out >> the room.
2:24:27
>> Wi-Fi is coming in a long time. What's up, boys? >> How you doing? >> What's going on? What's going on?
2:24:30
>> I couldn't bring the suit.
2:24:30
I thought the white coat was the next best.
2:24:33
>> No, no, white's great. The market's up. We're wearing white. It's very good. >> It's a good option. Okay.
2:24:36
So, >> break it down for us. What are you building?
2:24:39
>> Fin Twit was freaking out this morning because >> they uh they all thought they thought you paperclip them all. >> Okay.
2:24:47
>> Um so, yeah, talk about the launch.
2:24:47
So, uh, Nico was on the show a while back talking about kind of the pre-release.
2:24:53
Um, and the the app is live today.
2:24:55
Sounds like you've improved it quite a bit as well.
2:24:57
So, >> yeah, break it down. >> Yeah, sweet.
2:25:00
Thanks for having me, guys.
2:25:01
Um, >> yeah, Finwit's having a meltdown.
2:25:03
Um, they're having the kind of moment that we had August 2024 in software engineering.
2:25:08
Um, Carpathy tweeted, you know, cursor is kind of there now. >> Yeah. >> Um, >> yeah.
2:25:13
And what happened was there was a lot of hesitation from senior devs and I think a lot of the junior devs became really really good really really fast and there's a word for them now like context engineers and I think where we are now is there's this moment's happening for Excel.
2:25:23
So our product's called shortcut and it's a superhuman Excel agent and it does feel like that August moment for some things it's unbelievably better than humans and for others it's just like stupider than your intern. >> Yeah. Yeah.
2:25:34
>> Isn't the isn't the the dynamic of like the finance world and the Excel user radically different than the than the software engineer?
2:25:42
I feel like if you're if you're an amazing software engineer, you can work your way up at a big tech company, become like a senior engineering fellow, and basically always be working on algorithms and code and software your entire career and become fantastically wealthy, become an AI engineer that makes $100 million.
2:25:56
But in finance, you go to an investment bank or a hedge fund or a private equity shop, you're working in Excel, and then you level up and it becomes all about relationships.
2:26:06
And so you don't actually like when I think of like a managing director at you know KKR or something I don't think oh wow that guy's really good at XL I think that guy knows >> well he probably used to be probably exactly used to be.
2:26:18
So yeah how does this dynamic actually play out? >> Yeah.
2:26:22
So that's a good call out.
2:26:22
Um it's actually the same in tech but there's like two paths.
2:26:25
You could go you could go tech lead path and tech leads are like you know doing some wizardry on wherever algorithm or backend infrastructure.
2:26:32
You could go director path and it's just becomes like you talk about or you brag about how much headcount you have under you.
2:26:36
That is like the only thing that exists in finance.
2:26:40
>> And when we show people in finance this, their first reaction is like >> like it's kind of [ __ ] because like Excel is the reason that I'm where I'm at now.
2:26:47
And that path doesn't seem like it's going to be a thing. Um >> it is different.
2:26:52
>> I mean it feels like you'd like even I feel like a lot of people uh entrepreneurs don't want to be vibe coding themselves.
2:26:58
They want to hire somebody who's really good at vibe coding and they want to get the value of a senior engineer for the price of a junior engineer.
2:27:05
I imagine if I'm an MD and an investment bank and I'm like great uh I can even more quickly turn over a turn of the DCF to my client.
2:27:12
Uh and yes, I don't mind keeping the the the the Harvard grad, you know, up all night on the weekend, but if he could he or she could get it back to me in an hour instead of 10 hours, I'd be happy with that. >> Yeah. Yeah.
2:27:28
And you'd still rather that person work 30 hours a day and just send you more. Totally.
2:27:31
>> But yeah, we know this because of tech.
2:27:33
Like this happened and it was actually much scarier 2 years ago.
2:27:36
>> But whenever you can drive the cost of inputs lower and your ROI goes up, like the obvious thing to do is just put more gas on that fire. Totally.
2:27:42
Um it's just it makes sense that it's scary as in like we used to have tools and now we have things that can use our tools, right?
2:27:49
So um I think it's inevitable that they're going to learn what we learned and that's how I'm going to position it.
2:27:54
So, we need to go to the finance community and teach them about Jevans Paradox.
2:27:59
>> I I'm glad somebody's bringing this up.
2:28:01
>> We're going to host a Jeans Paradox happy hour.
2:28:02
Uh rooftop bar, >> you we're going to be we're going to be in New York.
2:28:06
I was about to say we'll start we'll just go on the street if we see someone dress a little circuit. Yeah, >> we should.
2:28:12
Are you Are you in New York?
2:28:13
>> No, we're in Menlo Park.
2:28:13
I'm the only person in Menlo Park who knows finance, unfortunately.
2:28:17
>> Hey, what about the venture firms? They they need to know.
2:28:19
Uh >> if you're serious by 100, >> you go to private equity or VC or private equity or hedge funds. >> Yeah, sure.
2:28:26
>> I was like, wait, what? >> No, no, no.
2:28:28
You put it in the Excel sheet.
2:28:30
Multiply the revenue by 100. That's the value.
2:28:32
>> You do a little bit of this and you close your eyes. >> That's the valuation.
2:28:37
>> Slap a multiply by 100.
2:28:39
>> Doesn't mean I need a model to to determine the valuation.
2:28:41
It's I'm just going to pay >> slightly more than my What is the shape of the user base?
2:28:46
of the user base? uh where is Excel actually like powerfully used because I feel like so many of the hedge funds they used to do stuff in Excel then it went to VBA then it went to you know high frequency trading and proprietary systems and I'm sure there's some stuff but it's usually in like private markets
2:29:01
less liquid markets like where where is XL holding on really strongly >> um you know I'd say hedge funds are like the toughest customers um anything that they're like really heavily relying on like already low-level systems or real-time data to public access markets like that is a tough customer. Thankfully, they're like a small sliver
2:29:17
Thankfully, they're like a small sliver and I'm not sure I'm building for them.
2:29:20
>> Um, but anything like private equity where you're getting a SIM and you need to build a model on it or a mini model, anything where like you have a template, you want to update it in light of new information.
2:29:26
Corporate real estate's like a bizarre area where there's a lot of PMF.
2:29:30
>> Um, but it seems to be that like building new models >> that be 100 times faster.
2:29:33
Now, >> editing existing models is a little tricky.
2:29:37
>> I mean, if you really want to swap a model for like new data, that's kind of like the highest bar that you can possibly have, but seems to be solvable. >> Sure.
2:29:44
>> So, what's the workflow now? I'm in PE. I have a SIM.
2:29:47
Uh, and for those that don't know, that's basically a bunch of PDFs and data showing like showing what's actually happening in a business that might be for sale. Can I drop how how soon? Let me walk you through.
2:30:00
Um, and that's exactly right.
2:30:02
So, it's like a confidential investment memo from, you know, the sell side.
2:30:05
A big bank will send it to a PE firm or they'll send it to a bunch of them.
2:30:08
Um, typically what you do is like you try to analyze, you diligence the deal and you look at this big PDF full of data and you try to model it to like understand is this a good deal, should I bring this to my boss or whatever.
2:30:18
You're usually constrained by how fast you could do that analysis.
2:30:21
Um, currently now you can just attach a PDF of a SIM.
2:30:23
They're usually PDFs or multiple PDFs or customer data dumps as well.
2:30:27
And you can just say like, hey, here's my template.
2:30:31
I use this for my LBO model. Um, make me 10 of them.
2:30:33
And in fact, now you can also just email it and be like, I want a hundred. Yeah.
2:30:37
>> Um and you can get all of that or you can attach a 100 attachments and ask for 100 models back and then it becomes more like how fast can I review data more like how then how quickly can I do ground work. >> Yeah.
2:30:46
Um what about just like FPNA orgs in like small and medium businesses?
2:30:49
I feel like when I've run companies I have like one finance person and we have a question uh should we expand into this market or should we add a new skew or should we do something?
2:31:00
Should we take out this line of credit or whatever?
2:31:02
And usually there's like an Excel model or Google Sheets model that's like built, but it's it's like not even there's not even like a canalist for pull a model off the shelf.
2:31:13
It's like no, I need to just be in Excel and model this out or even like CAC to LTV.
2:31:17
Is this how is how are my KPIs trending?
2:31:20
There's so much work that goes in the long tail of Excel.
2:31:23
Is that just like not the early market for you or do you see people abusing the tool?
2:31:30
>> I think the cool the crazy thing is is I do these live demos and people are like, "Oh, that's insane." But is this just DCFs? >> Yeah. Yeah. Yeah.
2:31:35
>> And I'm like, you know, we are in the tech world, so we think tech is the center of everything.
2:31:39
But anything you can possibly do in Excel is way more in distribution than like Ruby on Rails. >> Yeah. >> Right.
2:31:44
Like >> there's no there's no kind of model you can conceive that is like unfair to ask.
2:31:50
It becomes like what are the fundamental limits of AI?
2:31:51
And it turns out like writing Excel formulas is a lot easier than writing like amazing code actually.
2:31:58
So it's more more or less like how did this not exist at this point?
2:32:00
But yeah, I have to find >> you said uh you said shortcut beats first year analysts from McKenzie and Goldman head-to-head 89% when blindly judged by their managers.
2:32:09
We even gave humans 10x more time.
2:32:11
What what did that kind of like benchmarking exercise uh >> what did that look like?
2:32:19
>> We gave we gave multiple first year incoming analysts from Mackenzie Goldman, BCG, um JP Morgan, a couple other firms um like five tasks and we thought like they were the most in distribution for you know classic finance work.
2:32:30
So this was consulting like building op models, M&A models, LBO at DCF and like personal hobby stuff and actually even product management like build a dashboard over this customer dump.
2:32:39
Um we then gave them over 90 minutes per task.
2:32:42
Um some of the hobby ones had like 15 minutes and then we just asked them to submit it and then when we did that we also got in touch with managers from these firms and just sent them a Google form with like sideby sides completely anonymized and just pick your preference according to like accuracy, professionalism or whatever.
2:32:57
Um I thought it was going to be closer to 50/50.
2:32:58
I think what I really learned is first year analysts are like kind of lousy.
2:33:01
Like they're not that good at building Excel models.
2:33:03
Um and it's and then also I think people just take way longer to do things.
2:33:08
I think if I had to critique the study it would be that like five hours is not enough and like they would have liked to have a week one person wrote me.
2:33:15
>> Um but I didn't want to run that study. >> Yeah.
2:33:17
What's your read on the meter uh data that uh these coding agent tools did not actually speed up uh software engineers? Did you see this?
2:33:26
Uh so meter did a blinded study.
2:33:28
It was small small number but it was interesting because it was very solid software engineers working on advanced bugs in o in like big open source projects.
2:33:39
Not vibe coding a landing page like like truly like there is a sticky bug hanging out in some fundamental repo out there. Go fix it.
2:33:50
They estimated that they would be 20% faster, 30% faster, something like that.
2:33:54
Turns out they were 20% slower.
2:33:56
I'm wondering if you'll see a similar pattern where there's a speed up for bootstrapping and getting from zero to one on a project, but then for the really crazy person that already has all their macros and doesn't touch the mouse, maybe they're actually going to experience a slowdown.
2:34:11
Do you think that would happen?
2:34:12
What do you think about that? >> Yeah.
2:34:14
Yeah, it's a great question. I saw the study. I didn't read it deeply.
2:34:16
Um, what I will say >> would be bad for you.
2:34:19
It's just like it's just an interesting >> I think the obvious thing is that it's like using AI is a skill also, >> right?
2:34:25
So like the top startups have like the best context engineers in the world.
2:34:29
They know how to optimize the KV cache hit rate whatever it is.
2:34:32
>> Um if you kind of give like even a really strong engineer who's been coding for 20 years you're like now you have to use this tool or use it however you want.
2:34:38
>> Um I have no doubt that even the best engineers and the oldest engineers like eventually would find patterns in which it's massively unlocking for them.
2:34:44
>> So I think that is what's going to happen.
2:34:45
But I actually to be like the most critical I think clearly Excel modelers are way less technical and way less on the frontier of tech than our software engineers.
2:34:53
So it's actually more on me as like a product person to like how do I unearth some of these capabilities and hide the ones that they're not ready for um and even train people for this because it is 2023 all over again. >> Yeah.
2:35:03
How how templatable is the work?
2:35:05
Because I I know Canalyst sold to Teus which sold to Alpha sites I believe and canalyst was kind of like off-the-shelf financial models some data integrations.
2:35:15
Capital IQ has had integrations.
2:35:15
You have a Bloomberg terminal you can have integrations and so there's almost a world where you want to like sit on top of a of a library of templates.
2:35:23
In some ways, the way like deep research is clearly RLHDF on like a couple templates of like what a report looks like. It loves tables.
2:35:34
And so I I don't know if you want to do like fine-tuning or like have a model that selects a pre pre-filled template and then you're using your tool just to update that template.
2:35:45
What do you think about that? >> Yeah.
2:35:47
So let's break it down into two things.
2:35:48
One would be like product and then the other one would be training.
2:35:50
Um as far as product is go as far as product is concerned like I think the most important thing you can do is let users actually upload their templates like your income statement is going to be different than you know or pinkus or whatever um and then just let them work naturally from theirs but then from the training perspective you have to collect as much of this data as possible and train your models.
2:36:07
Um kind of definitionally if you're using frontier models you can be limited in what you can and cannot serve in terms of trained models.
2:36:14
Um but it's it's actually a much more constrained environment.
2:36:16
So like I think we stand to benefit benefit from it much more than does like operator or you know deep research. >> Yeah. Yeah. Makes sense. Jordy, anything else?
2:36:24
>> Um where how do you think the the product uh needs to improve going forward?
2:36:29
It sounds like in the in the study uh in the test that you ran with the the incoming analysts, it performed very well.
2:36:36
Uh that's kind of like a a unique situation.
2:36:39
What's it going to take to to get it to the point where it's actually >> disrupting the job market in you know in in in some of those roles?
2:36:49
>> Yeah, it's a great question.
2:36:49
Um, clearly right now I think what it does is it makes like maybe you guys are rustier on Excel that you used to be.
2:36:55
Like it makes you much better at Excel and if you're kind of a rookie or like you're a solo like prneur and you want to like have a oneman finance shop like you're much better now already with it.
2:37:03
So for that like we passed the capabilities threshold.
2:37:06
The big question becomes like for like real enterprises who people hardcore use it, are we there?
2:37:09
And I would say like if August is that moment, we're like in June.
2:37:13
Like it is a pretty clear capability threshold we have to pass and we're actually actively adding all of the limitations to our internal benchmark and just hill climbing them as aggressively as we can.
2:37:22
Um part of it becomes a little bit of a product science of like how you can abstract the limitations away from them.
2:37:27
Um, but like more specifically, it's exactly editing and overwriting large nasty existing templates and files with new data.
2:37:36
It's like, yeah, if you don't do that at a certain percentage of accuracy, it's just a non-starter.
2:37:42
>> Yeah, I feel like part of the challenge will be figuring out like giving the user tools to figure out where shortcut is hallucinating, right?
2:37:49
is hallucinating, right? if you have like part of that I I I remember you know working with with um you know I'll get a model made in the past >> and I just look at it and I'm like okay everything looks good but like it's definitely off and like it's just like it's too good or it's too bad
2:38:07
>> bad model smell >> yeah exactly um and then you dive into it and you find like one or one or two kind of reasons reasons for that um and I think like the pe people that might critique Shortcut today are going to say like, "Oh, I used it for this or that and like it missed this column or whatever, you know, um, that kind of stuff." >> So, I mean, in terms of how we think
2:38:29
>> So, I mean, in terms of how we think about it, I'll tell you how we think about it and then like what actively we're doing about it, but >> it's like this is not a new phenomenon like in coding it had to get certain
2:38:37
good and it had to be certainly like at a certain level of observability and when cursor like brought the apply diff function, it actually made the job of supervising AI like doable like I used to copy blocks and it would be I wouldn't know what would break. Um, that
2:38:47
Um, that was the big thing.
2:38:49
was actually observability, not accuracy.
2:38:49
And then Sonic got good enough.
2:38:51
Um, in like self-driving, it's safer than we are, but we still don't want to make that trade because we want to be like that guy killed that person, right?
2:38:56
Like we need to be able to blame people.
2:38:58
Um, radiology, it's there too, right?
2:38:59
Like we just don't know who to sue if something went wrong.
2:39:03
Um, I think for finance, like exactly what you're talking about is we had to have a UI that allows for you to observe the diff.
2:39:10
Um, and like we had a good inspiration, right, cursor.
2:39:11
So actually now it's part of the launch video I just I just shared.
2:39:14
Um, and I know we were talking about like product demos, which I have a strong like, you know, opinion about.
2:39:20
Um, like the whole like you have to be able to see everything that's changed, but also the root material because in Excel definitionally some things are hardcoded, but you want to know where those came from because you can't justify it, right?
2:39:29
Like where what part of the 10K did it come from?
2:39:30
What web search did it come from? What's the exact quote? Like what page?
2:39:33
So if I I really do believe that it's not about getting to 99. 9% accuracy.
2:39:36
It's about like getting to perfect traceability cuz like you know your analysts suck.
2:39:41
Like even your associates are not good, right?
2:39:44
So you just have to be able to to observe it at a superhuman level.
2:39:48
>> Do you think >> Well, yeah.
2:39:48
And the the good thing for analysts is like you combined with a product like this could actually get to the point where you're truly elite. >> Yeah.
2:39:56
So I know people are afraid. Let me hit that.
2:39:58
Like clearly analysts should be using this or a tool that will try to do what we're doing. Um no doubt about it.
2:40:04
It's >> the right launch is a non-zero amount of controversy, right?
2:40:08
Like I know I know that's an emotional reaction that people have, but clearly people will be using this the same way.
2:40:13
Like if you don't use cursor now, like you don't know what you're doing, right? >> Yeah. Yeah. >> Yeah.
2:40:17
Do you think >> you're not afraid to trigger trigger a couple of incoming analysts that are say I used it and and it did this wrong.
2:40:25
It's like well you probably would have made a mistake too. Yeah.
2:40:27
Do you think you'll face more competition or that there will be more war rooms planning to compete with you from the hyperscalers who have products that they want to add this feature to or the foundation model labs who see hey uh you know what I can maybe I don't have control over a consumer application but I can do a lot of these calculations in pandas in Python.
2:40:54
>> Yeah I'll answer those.
2:40:56
>> Yeah those two are very different. >> Yeah.
2:40:57
Um, when it comes to the hyperscaler, it comes down to can they? >> Mhm.
2:41:00
>> Um, I think they would if they could, right?
2:41:03
And they would have already.
2:41:03
Um, I think it comes down to like inertia, talent even.
2:41:08
>> Um, then that's not true for open AI, right?
2:41:11
>> Uh, but what's what's the problem there?
2:41:13
I think >> it's going to be a bitter competition is the truth.
2:41:16
Um, >> then you have to ask what principles do you really believe in and will it's like prevent you from doing what I'm doing, >> right?
2:41:22
Like I think even if you made an effort to copy me or beat me or directly compete with me, you'd have to believe the things I believe in like wholeheartedly.
2:41:27
And I think most of the people at Frontier Labs I know cuz I work at a small one like really don't even care about Excel or they think that like we're all going towards an input output model anyway. >> Sure. >> Right.
2:41:37
That like you have to do a general training like paradigm to bring you just a model out at the very end and who cares if it's hardcoded like is a thing of the past anyway.
2:41:44
Um I think there's like a bajillion dollars to make in between those two states.
2:41:48
Y >> um and I'm not even sure that their path brings you to that state first, right?
2:41:54
Um so my argument or the way I think about competition is more like in principle versus the top labs, but it's more just maybe arrogance when it comes to hyperscalers.
2:42:05
>> No, no, I no I I think that makes a lot of sense.
2:42:07
Um I mean OpenAI has is fighting a war on like seven different fronts right now.
2:42:12
They're like, "Oh, we're also going to do a phone.
2:42:13
We're also going to do a browser and we're also going to do this and that and you know like a coding environment and all this and like and that comes from the research.
2:42:22
>> How do you Yeah, I guess you're you're seven steps down the power law.
2:42:27
>> Will they really build a competitor a piece of spreadsheet software like they probably could but that's pretty low on the priority stack for them I imagine.
2:42:34
I guess I guess h how your approach from a go to market standpoint just seems to be like create the best product possible and release it and see what happens.
2:42:45
>> Is that will you have an enterprise motion over time?
2:42:48
Do you know are you going to set up a New York office and like pound the pavement? What? What?
2:42:53
>> No, you we're going for drinks, guys.
2:42:55
We're like Now, now listen.
2:42:55
Um I think people have a lot of false stupid pride about like their go to market motion, right?
2:43:02
Like I I've mentioned cursive four times, a lot of admiration, but they like they beat the drum that like we've never spent the dollar on go to market or distribution or ads.
2:43:08
Like >> that's a flex, but it's stupid, right?
2:43:12
Like you can also have a world-class sales team.
2:43:14
>> Um when I meet with the CIOS of the biggest banks in the world, like they want to go right now.
2:43:18
Now I need to know like what it takes to have the right SWAT team to put that together, but I don't think that comes at the expense of what we're doing from like a proumer bottoms up path. Yeah, totally.
2:43:28
>> Inevitably, one will seem to be more fruitful than the other and I will >> pour proportional resources towards it.
2:43:33
I think clearly my gift is more bottoms up.
2:43:36
>> Um but both be positive some for you >> for sure.
2:43:39
And I don't think Windsurf had Yeah, exactly. >> Exactly. Exactly. presence.
2:43:45
>> Uh, deal director in the chat says, "Nico said the finance bro end of times is delayed a couple months," which I think is correct.
2:43:52
>> I wanted to give them a chance to get back.
2:43:55
>> Yeah, you have two months to escape the permanent underclass. >> Yes. Yes. Yes.
2:43:58
So, start vibe vibe modeling. >> Anyway, fantastic.
2:44:01
Always good hanging out with you. Jordy, anything else? >> Yeah. Congrats on the launch.
2:44:06
>> Yeah, congrats on the launch.
2:44:07
>> We'll talk to you soon. >> Appreciate it, guys. See you. See you next time. This You're on notice.
2:44:09
I'm a I'm a fixture now, but I'll actually suit up. >> Okay. Okay. Yeah.
2:44:14
The microphone's fantastic.
2:44:16
Everything else is fantastic. >> Take my job serious. >> Yeah. >> Bye.
2:44:20
>> Gary Tan says, "Hire the engineer who worked on the best product.
2:44:22
Hire the marketer and salesperson who managed to get the worst product to out sell the best product."
2:44:30
>> I I'm I'm going to >> You agree?
2:44:32
>> So, I mean, an example here is is um uh this is the the product quality is up for debate.
2:44:38
I'd say like the market generally >> I know where you're going with this. Yeah.
2:44:42
Wind serve team >> but they got to 80 million.
2:44:45
>> They had a fantastic go to market motion and now that is a cognition >> and now they had a cracks engineering team and >> pretty yeah pretty pretty wild. Yeah.
2:44:51
I mean it is it is hard to find I like searching for engineers who worked at the best products. It's very easy.
2:44:58
It's what Mark Zuckerberg is doing.
2:44:59
Go hire the the open AI researchers who built 03.
2:45:03
Go hire the open AI researchers who worked on images and chat GPT. They're clearly good.
2:45:07
It's much harder to actually identify products that have outsiz distribution, right?
2:45:13
>> Funny that OpenAI has so many different like products and models and things like that that there's so many people that can leave and be like, I was the product lead on >> 03 Pro Mini High and you can raise $100 million off that.
2:45:26
>> You can get $100 million in your bank account in a uh in a hiring bonus.
2:45:28
Uh Noah Smith says, "I was doing a podcast with Eric Torberg and Darwh the other day and we were talking about whether AI will destroy humanity and in an off-hand way I remarked that technology had already destroyed humanity and it's showing the world population with a projection based on fertility rate of 1.
2:45:47
6 children per women basically showing that our by 2800 our population will go down to zero.
2:45:53
I would just say uh technology has not destroyed >> 775 years to figure out this problem. >> Skill issue. >> Skill issue.
2:46:04
>> I I think we're gonna be good.
2:46:04
But uh yeah, it it is it is an interesting trend.
2:46:08
And obviously uh it it's one of the less explored territories in what happens.
2:46:14
>> I think if we can get a a a vaccine for twins and triplets >> every and we just keep >> you know what that is?
2:46:21
It's not being >> keep the pregnancy rate where it is.
2:46:24
It's it's it's uh the rate of twins and triplets is very low in vegetarians and so eating a carnivore diet might literally save us.
2:46:33
I'm not kidding about this.
2:46:34
Uh >> governmentrun hamburger shops. >> Yes. Yes. Yes.
2:46:39
>> Every universal basic hamburger three times a day.
2:46:41
times a day. It is it is one of the unexplored answers and I think because tech people debate the what happened in 1971 so much and the stagnation thesis no one really wants to interrogate was it technology that happened in 1971 was that the beginning of the technology
2:46:56
boom but maybe we should be having that conversation about the effect that you know the personal computer and then the internet had on everything that might be one of the reasons uh one of the interesting uh theories for you know the original what happened in 1971 the peter teal stagnation thesis. This idea that
2:47:12
This idea that we are not building flying cars, we stop building nuclear reactors is truly just that the internet was such a great draw that it brain drained everyone away from everything because you over in nuclear territory, you had regulation and you couldn't move that fast.
2:47:28
But on the internet, you could throw up a a website and get distribution and had very few uh very few few forms to fill out, very few gatekeepers to delay your uh your rise.
2:47:39
And so you could just grow at the natural rate uh that the that the market pulled.
2:47:43
You basically had a free market on the internet and in the computing world and you did not have a free market in the energy world or in the flying car world.
2:47:52
And so uh it is it is interesting.
2:47:56
>> Well, more importantly, uh armorplated golf force one golf force one spotted with President Trump at the golf course.
2:48:03
I guess did he get this transported to Scotland? >> I suppose so.
2:48:07
>> It looks like a Lynx course.
2:48:07
I mean, they fly on a on a global like, have you seen the plane that the president flies on when they bring the beast and the SUVs and stuff?
2:48:16
>> Oh, they bring they bring uh is it the C130?
2:48:19
Yeah, like it's one of those huge tube transport and then they just I think the entire front of the or maybe the back of the plane goes down.
2:48:26
They just they're already in the cars and they just drive out in the armorplated SUVs.
2:48:32
And so bringing this does not seem that difficult.
2:48:34
It is kind of funny because um I mean I guess it keeps you safe while you're in it, but then you have to get out to golf.
2:48:39
So So it's not like a full security detail, but um seems like it works to some degree.
2:48:44
Uh anyway, Nizzy says, "The only way to make AI output good code, I swear, is copy the Figma file." Exactly. No mistakes.
2:48:53
My sister will die if you f up. Ridiculous.
2:48:56
Um, and so, uh, yeah, this is, uh, you know, I I I think this is more of a joke at this point and like we are out of the like make a threat.
2:49:06
>> If it works, if it works, it works.
2:49:07
>> I don't think I don't think it actually I don't think that I don't think that you actually get better LLM results by prompt engineering in that way.
2:49:13
I think that's a I think that's a relic of a few years ago where you used to have to say, you know, uh, please do this and don't make mistakes and all of that stuff.
2:49:22
Like you used to when you went to prompt an image generator, you used to have to say no six fingers. Don't do six fingers.
2:49:29
And then it would do five fingers. You had to tell it that.
2:49:30
And now it's all baked in.
2:49:31
And so you don't run into that problem at all anymore.
2:49:35
Anyway, in other news, Ashley Vance has a new video on on core memory about new limit, who's been on the show, uh, which is finding combinations of protein that reverse aging across the body.
2:49:44
Is perhaps the most exciting work in the biotech field. I agree.
2:49:47
It's backed by Brian Armstrong, Patrick Collison, Joshua Kushner, Nat Freriedman, and others.
2:49:52
You should go check it out. It's a 17minute video. It's out on X now.
2:49:54
And of course, it's on the core memory YouTube. >> Yeah. Shout out to Ash.
2:50:01
>> I mean, Ashley has got to do a video about Robert Maddox. >> Robert Maddox. Yeah.
2:50:05
>> Begging for a core memory deep dive.
2:50:07
>> This feels equally important to humanity's future. >> Yes. Yes.
2:50:09
So, Robert Maddox has a hobby.
2:50:12
Uh, looks like an older gentleman.
2:50:12
He h his hobby is installing jet engines on anything that can move.
2:50:18
And he put a jet engine on a bicycle here and is just ripping down the street. Uh absolute legend.
2:50:28
>> Let's put a jet engine on a horse.
2:50:30
>> Let's put a jet engine.
2:50:31
>> Let's put a jet engine on a on the basketball player.
2:50:34
I don't even know his name.
2:50:35
The guy the guy who has the horse that that he >> Yeah. Yeah. I don't know. >> Someone else.
2:50:41
>> I don't even This >> Nicola Joic. Joke kick.
2:50:42
I'm sure a jet engine feeling my non non my nelson. >> Look at this. Look at this. Look at this. Here we go. Look at this guy.
2:50:51
He's having the time of his life.
2:50:53
Escape riding riding a jet engine.
2:50:54
I didn't know you could I wonder what it takes to build a jet engine like that. Yeah.
2:50:58
The metal just heats up like that. It's crazy, right? >> That's insane. >> Wow. Yeah. This is hard tech. Get this into the guno. >> Wow.
2:51:05
Santa, >> this is remarkable. >> I wonder what a rig.
2:51:08
>> Why is this not like commercialized?
2:51:11
Like why is this not a thing?
2:51:11
Is it just extremely dangerous?
2:51:12
Did they never figure out how to make jet engines reliable in the right way?
2:51:15
It feels like maybe it's not fuel efficient. I don't know. We use them in planes. Why not in skateboards?
2:51:22
This see this seems extremely dangerous.
2:51:24
I wonder how fast he's going.
2:51:25
The Santa Claus ones particularly seemingly does these in flat open space.
2:51:30
>> This also feels like textbook like this could be AI generated, >> but I don't think it's actually Wait, is this real?
2:51:35
I thought it was AI when I saw it. >> I don't know. >> It's farm to table.
2:51:38
It screams Robert Maddox. Let's look it up.
2:51:40
Let's Let's fact check this.
2:51:43
>> He has an account called Crazy Rocket Man.
2:51:46
>> I think he's I think he's Lindy.
2:51:46
I think he was around before Generative AI. >> I can see videos. >> Is real.
2:51:51
>> He was doing this 10 years ago. >> 10 years ago.
2:51:52
You think you think he was using Dolly 1. 0 for that? No way. >> No way. >> No way.
2:51:58
This is >> Or he's been an entirely Sam a figment of Sam Alman's imagination. >> Potentially.
2:52:05
>> Okay, we're getting back into 20 years ago. He's 17 years ago. aticsjets. com.
2:52:12
>> Incredibly scary pulset powered jet engine car. >> There we go.
2:52:16
Well, speaking of AI, Elon Musk has rolled out a change to super Grock coming to your phone soon.
2:52:21
And this is actually >> super great feature.
2:52:24
So, it has an auto setting that chooses the best mode whether to go with the fast response or the deep or the expert response.
2:52:32
the deep or the expert response. This is something I it's turning into one of my number one frustrations with the chat GPT app at the moment is that sometimes I want 04 and I want or 40 and I want something quick and then sometimes I
2:52:44
want 03 and if I trigger the quick question with 03 pro and it's waiting there for 10 minutes I have to go and start a new chat and then change the model and then get the o the the the 40 question answered and it just slows me down a little bit and I think this auto mode should come to the chat box. I
2:53:00
I would even just be down for give me the ability to pick the model in the chat box.
2:53:07
So I should just be able to say hey use 40 to tell me the capital of Scotland and it should just do it quickly or I should say I need a deep research report on this and it should just trigger deep research.
2:53:19
I shouldn't have to click UI buttons on this text is the universal interface.
2:53:23
You know who said that OpenAI.
2:53:25
So, make it truly a universal interface by allowing me to pick the model in the prompt window, please.
2:53:32
Um, anyway, speaking of chatbt, Quinn Nelson has a post that says, "Today I learned my wife is using Chat GPT to find coupon codes for web stores online.
2:53:42
She said it works almost every time without adridden websites filled with old codes.
2:53:46
I've literally never once considered using an LLM for this, but now it seems obvious. I love it." >> Uh, very interesting.
2:53:54
I would never I I never would have thought to to do this, but now that it's out there, uh, makes a ton of sense.
2:54:00
Go when you're searching, when you're about to check out, flip over to >> check your checked Lucy, code Ben 20 gives you 20% off storewide. >> There we go.
2:54:10
>> So, >> uh, it's uh, this is breaking news and this is important.
2:54:15
Mercedes-Benz is finally integrating Microsoft Teams into its vehicles.
2:54:19
We have been begging for this. Yes, >> they really are.
2:54:22
You know, it's basically an office on wheels. This is about time.
2:54:27
>> This is huge because uh if you're a defense tech founder, you're probably ITAR compliant.
2:54:31
You need Microsoft Teams.
2:54:32
You're probably driving a Mercedes-Benz 6x6. >> Yeah. >> Right.
2:54:36
>> You should be >> G63 6x6. >> Not very American.
2:54:38
>> Probably armor plated.
2:54:38
That's extremely Yeah, it's pretty pretty American dynamism.
2:54:45
>> You're making getting an Americanmade 6x6.
2:54:48
>> I mean, there just aren't that many other options.
2:54:49
If you're if you're in the market for a 6x teams in a Ford Raptor.
2:54:52
That's the That's the ultimate goal.
2:54:55
>> The gauntlet has been thrown down by Mercedes-Benz. Balls in your court. Ford Motor Company. Let's get >> So funny.
2:55:00
Everybody's dunking on this being like, "Oh, this is so silly. No one asked for this."
2:55:05
It's like, how so many people take conference calls while driving?
2:55:10
Why would they not make an integration?
2:55:13
>> Also, Mercedes-Benz has one of the few active systems on the road that's level three self-driving.
2:55:19
three self-driving. And what that means is that when you're on the freeway at certain speeds, I believe between like 20 and 40 miles an hour, they're so they are so confident in their self-driving technology that you can watch a YouTube video while you're in the driver's seat,
2:55:36
which doesn't sound that crazy when you think about like Whimo and what Tesla's doing with a robo taxi, but it is an interesting thing for just like a consumer car that you can actually own and you can drive around and then you can throw it into self-driving mode and it will let you watch a YouTube video. So, it should let you take a team's
2:55:53
So, it should let you take a team's call, I would imagine, or let you use anything.
2:55:56
And then it disables the screen if it needs you to take over.
2:55:57
And I think the reason that they've rolled out level three at like low speeds in traffic is because even if you get into an accident at 40 miles hour, if it's a new Mercedes-Benz, it's probably very, very hard for that accident to be truly catastrophic because you're just not going that fast. Um, but uh I don't know.
2:56:16
We'll see how this works out.
2:56:18
We'll see if the team's users enjoy it.
2:56:20
Anyway, >> uh in more news, uh Sebass in the chat says, "What did you think of the Tesla Samsung deal?" Which I didn't see. >> I didn't see that. >> So, I pulled it up. >> Okay.
2:56:31
>> Um this is this is a bad look on our part, John.
2:56:34
How do How do we miss a $16.
2:56:35
5 billion supply deal between Tesla and Samsung?
2:56:40
>> To spur the chipmaker US contract business. >> Okay. Interesting.
2:56:44
So Tesla has signed a $16.
2:56:45
5 billion deal to source chips from Samsung. Yep.
2:56:47
A move that could bolster the South Korean tech giant's unprofitable contract business, but is unlikely to help Tesla sell more EVs or roll robo taxis more quickly.
2:56:57
Tesla CEO uh Elon of course said Sunday that Samsung's new chip factory in Taylor, Texas would make Tesla's next generation A16 chip.
2:57:07
And uh so anyways, we'll need to um look into this more.
2:57:11
It says production is still years away. >> Okay.
2:57:16
>> So it's unlikely to help Tesla's immediate challenges.
2:57:19
>> It's still just notable that they're not making the chips that go into Tesla's self-driving units in at TSMC.
2:57:24
self-driving units in at TSMC. like that that is what's the the real news here because with everything on the cutting edge of AI whether it's Apple intelligence chat GPT Google Gemini no matter what the chip is it's fab TSMC
2:57:40
right now on the most on the most leading on the most leading edge >> process so and interestingly um for a long time TSMC's most cutting edge uh fabs were driven by by smartphone demand So Apple would come to TSMC and say we need the the smallest nanometer fab you have. We want 2 nanometer. We want 3 We want 2 nanometer. We want 3 nanometer.
2:58:06
We want the most cutting edge.
2:58:07
And the most cutting edge chip, the most cutting edge process would always go into a smartphone.
2:58:11
And the reason for that was not raw compute power.
2:58:15
It was compute efficiency and power efficiency.
2:58:17
Because if you're putting a chip in a phone, battery life is really, really key.
2:58:20
If you're putting a chip in a data center, power efficiency is important on the big macro scale, but at the end of the day, if you're getting a one gigawatt data center with a nuclear power plant and a natural gas peaker plant outside and you have solar panels everywhere and batteries, you have all this stuff.
2:58:38
Like power is important, but it's not as critical.
2:58:43
Like you can just pay for it, whereas you really can't just pay for more battery life magically in the iPhone.
2:58:52
and it certain TSMC has always been driven to the leading edge by Apple and by the smartphone makers but that's actually shifting now and so >> earnings happened last week and they blew them out and >> yeah so Samsung is up over 6% today in reaction to the news somebody was commenting about the news uh Jesse Pelton most important news of the year so far is what Jesse says Elon says, "So few understand this." Few understand this. >> Few >> few. >> Uh, no.
2:59:29
He says, "Elon says it will become obvious in two to three years why this is big news."
2:59:32
>> Well, yeah, because I mean, I would imagine that what they're going to fab is something that looks like a AI ASIC for something like the TPU, but but designed specifically for Tesla's self-driving capabilities, which might be different than a LLM.
2:59:45
In >> America, made in Texas.
2:59:47
you when you are when you are training a self-driving system.
2:59:52
You you don't need to know all of humanity's knowledge forever necessarily.
2:59:57
You might not need to train on all of the web text.
2:59:59
You need a ton of images and it might be a slightly different pipeline.
3:00:02
And so optimizing around that might be what they're thinking of doing.
3:00:06
We we'll need to dig in more.
3:00:08
in more. But I would see this as similar to like TPU or tranium or any of these other uh Apple like the Apple silicon chips like it is they are a magsaving company that has slightly different needs for their particular compute demands and so they are going to one of
3:00:27
the two or three leading fabs and and making exactly what they want at scale and they're planning to ship a lot of these because they you know what I think what Elon is saying is like we're we're betting on needing a ton of these chips because we're not planning to sell less cars. We're going to sell a lot of these
3:00:44
We're going to sell a lot of these cars and they're all going to have custom chips in them.
3:00:49
>> And this will be another edge for Tesla on autonomous driving. >> Exactly.
3:00:53
Especially once they add a V12 >> with straight piped exhaust that >> naturally aspirated.
3:00:58
>> Adding that back will actually be easier than than than the Ford Raptor getting custom silicon from from uh from Samsung. That's the bull case.
3:01:06
Anyway, >> well, thank you for tuning us tuning uh into TVPN today.
3:01:11
We have to get on with London right now.
3:01:16
>> Uh but uh we appreciate >> us five stars on Apple podcast.
3:01:17
Leave us five stars on Spotify.
3:01:19
We are the number four ranked podcast in technology on Spotify.
3:01:24
Thank you to everyone who supported us to help us get here. It's been a wild ride. We enjoy so much. We appreciate you.
3:01:32
Especially everyone in the chat.
3:01:34
Especially Huxley and Jonathan >> uh the deal director >> and everyone. >> Everybody.
3:01:40
>> Thank you for watching.
3:01:40
We will see you tomorrow. Have a great day. >> Have a great evening. Goodbye. Cheers.