Latest / In the Loop AI / AI NEWS: Meta's $25B Fumble | Anthropic $200B Valuation | OpenAI Acquires Statsig

Transcript

AI NEWS: Meta's $25B Fumble | Anthropic $200B Valuation | OpenAI Acquires Statsig · In the Loop AI

0:03
Ejaaz: Okay, the AI talent wars are back and Meta is bleeding after spending $25 billion
0:10
Ejaaz: to acquire the best AI researchers ever.
0:13
Ejaaz: They just lost eight of them in the last week. And Josh, Sam Altman is laughing and fighting back.
0:20
Ejaaz: He just spent $1.1 billion to acquire a new startup and some of the best AI researchers.
0:25
Ejaaz: And Anthropic quietly raised another casual $13 billion, which now values them at $200 billion.
0:32
Ejaaz: And I heard some whispers on the grapevine that Apple might be making a comeback in AI.
0:38
Ejaaz: Folks, we have got a jam-packed week of news updates, and we're going to get right into it.
0:44
Ejaaz: But let's start with Meta Superintelligence Labs, which, if you remember,
0:48
Ejaaz: was the new superhuman AI core unit that Zuckerberg set up and actually spent
0:54
Ejaaz: a lot of time poaching some of the best people from OpenAI, Anthropic,
0:59
Ejaaz: and investing in this startup called ScaleAI around $15 billion to form this
1:04
Ejaaz: kind of like Avengers of AI.
1:06
Ejaaz: And the hope was that Meta, who has consistently been dragging behind on AI
1:10
Ejaaz: models, will jump and leapfrog to the top with a new AI model that is better
1:15
Ejaaz: than ChatGPT and whatever you might say.
1:17
Ejaaz: But within two weeks, Josh, of them forming this team, eight executives,
1:22
Ejaaz: you heard that correctly, Eight executives have left with these guys highlighted
1:26
Ejaaz: over here being some of the main ones.
1:28
Ejaaz: And if you were thinking, maybe these are just some casual engineers or people
1:32
Ejaaz: that got caught in kind of like
1:33
Ejaaz: the org shift from the old meta AI team to the new one, you'd be wrong.
1:38
Ejaaz: These are the executives, some of the leaders that were appointed to this new team.
1:45
Ejaaz: It doesn't signal the strongest amount of confidence.
1:48
Ejaaz: I remember when we were talking about Zuck taking this aggressive move in the
1:52
Ejaaz: AI talent wars, you know, three weeks ago.
1:55
Ejaaz: We were very confident that Zuck is making the right move here.
1:58
Ejaaz: He's done kind of similar moves in the past when he acquired WhatsApp,
2:02
Ejaaz: when he acquired the Instagram team.
2:03
Ejaaz: And we kind of thought this was going to be the same type of move as well.
2:07
Ejaaz: But this doesn't inspire confidence to me, at least. Josh, do you maybe have
2:11
Ejaaz: another take or do you feel the same?
2:13
Josh: Yeah this seems like things are getting messy i mean when you move this quick
2:17
Josh: i mean it's funny back in the day uh mark zuckerberg and facebook's ethos was
2:21
Josh: move fast and break things that was the theme of their company that was their
2:24
Josh: motto and that's exactly what they're doing now i know that wasn't always the
2:28
Josh: plan but that seems to be the case that what they're doing now actually if you
2:30
Josh: can go back to that list you guys i just want to kind of walk through a few of those members
2:33
Josh: it's important to note that some of these people have been around for a very
2:36
Josh: long time so one of them worked with meta for more than eight years spent five
2:40
Josh: years at meta uh eight years at meta
2:43
Josh: a lot of since 2024 so a lot of these people have been there
2:45
Josh: for a long period of time these are the executives that have been
2:48
Josh: around that have been building with the company presumably very happy with the
2:51
Josh: company who are now leaving as this new order comes in so i'm not sure we could
2:56
Josh: view this as totally bearish because this very much feels like an instance where
3:00
Josh: the old guard is just kind of like not really interested in this new mission
3:05
Josh: as they pivot as a company as they bring in all this new talent
3:08
Josh: What I would be concerned about is if the new people who are recently hired
3:12
Josh: weren't doing well and wanting to leave because that's where I would start throwing
3:16
Josh: red flags. I'm like, OK, old people make sense. They've been around.
3:18
Josh: They've probably secured more than enough money to take care of themselves and
3:21
Josh: their families. And they're ready to move on and let this new like young blood come into the company.
3:25
Ejaaz: Well, that's actually what's happening, Josh. If you look here,
3:29
Ejaaz: one of the senior vice presidents of Scale AI,
3:32
Ejaaz: which was the company that Zuck basically spent $15 billion to acquire their
3:37
Ejaaz: talent, I think less than a month ago, left, right?
3:41
Ejaaz: And a few other executives also left. tenant ones
3:44
Ejaaz: it's a huge red flag and i was kind of thinking
3:47
Ejaaz: about like why this might be the case like you know
3:50
Ejaaz: you've just signed 100 to 300 million dollar
3:54
Ejaaz: um bonus packages um you're set to probably make a couple billion but for staying
3:59
Ejaaz: at meta and at least giving it a shot maybe give it more than two weeks right
4:03
Ejaaz: and then i was thinking well maybe they just kind of took the 100 million sign
4:06
Ejaaz: on bonus and then quit um but chi hao wu who is a former ai and ml specialist with Meta,
4:13
Ejaaz: he was kind of like interviewed about why this might be happening.
4:16
Ejaaz: And he said, speaking generally and not for myself, a lot of people in the AI
4:19
Ejaaz: team maybe feel things are too dynamic.
4:22
Ejaaz: There were a lot of organizational changes. And in fact, my manager was changed
4:27
Ejaaz: several times. So we might just be seeing kind of like the effects of churn
4:31
Ejaaz: happening throughout the team and the organizational movement.
4:35
Ejaaz: And maybe things settle down over the next couple of months. I don't know.
4:38
Josh: It seems like this is probably too early to judge the success of these decisions.
4:42
Josh: I mean, he spends a lot of money. I assume it's going to ruffle a lot of feathers
4:45
Josh: internally as well as externally.
4:47
Josh: And they're just kind of in a period where they're figuring out this new structure
4:51
Josh: because there's a whole new leadership hierarchy, right?
4:53
Josh: I mean, they hired Alexander Wang here and they brought him over and it seems
4:57
Josh: like it's going well, but apparently it's not.
4:59
Josh: They may have overpaid, which might be like a gross unjustification because
5:03
Josh: like clearly they overpaid, but how much did they overpay? I guess is my question I have to you, Ejaz.
5:07
Ejaaz: I just want to say when Zuck originally was announced to spend this amount of
5:12
Ejaaz: money on Scala and Alexander Wang specifically.
5:15
Ejaaz: There was a massive question mark above that, right? Now, we've seen a lot of
5:20
Ejaaz: crazy stuff happen in AI investment.
5:22
Ejaaz: We've seen billions being spent on compute and video chips and all that kind of stuff.
5:26
Ejaaz: And this was kind of like the next step in that directional trend, right?
5:31
Ejaaz: Is it worth spending $15 billion on an AI data company? More in so,
5:35
Ejaaz: is it worth spending that much on one person?
5:39
Ejaaz: And people thought, well, Zuck's been right before. Maybe he'll be right again.
5:42
Ejaaz: But the criticism that we're seeing now that some of these key executives from
5:46
Ejaaz: the exact company is leaving is is causing a lot of worry.
5:50
Ejaaz: And Alexander Wang is kind of like caught in the crossfire between all of this happening.
5:54
Ejaaz: You'll see like one of the points being made on this tweet over here is that,
5:58
Ejaaz: you know, Surge and Mercore, which is a reference to Scale AI's competitors,
6:04
Ejaaz: are doing as good, if not better, a job than Scale AI. You know,
6:07
Ejaaz: they do pretty much the same thing.
6:09
Ejaaz: And I'll value that only a fraction of what scale AI is. So the question now
6:13
Ejaaz: becomes, well, maybe has Zuck overpaid?
6:16
Ejaaz: I'm kind of thinking yes right now. But again, it's too early to call.
6:20
Ejaaz: We have to see what they actually produce. I say give it a couple of months.
6:24
Josh: Okay, this is what I was going to ask. Do you still feel confident or do you
6:27
Josh: not feel confident in Meta's ability, in Zuck's ability to turn this around,
6:31
Josh: to become a real competitor in the AI space to the likes of like Anthropic,
6:34
Josh: OpenAI, XAI, the whole crew?
6:36
Ejaaz: Okay, I'll tell you what my gut is telling me. Short term, yes.
6:42
Ejaaz: But long term, I'm undecided. And the only reason behind that is Zuck hasn't
6:49
Ejaaz: really shown his proficiency in infrastructure before.
6:52
Ejaaz: He's a consumer guy, you know, he's made some of the best consumer apps,
6:55
Ejaaz: consumer experiences that people have ever seen.
6:57
Ejaaz: And I'm confident in him translating that towards AI apps.
7:01
Ejaaz: I just don't know whether he can kind of like operate that at the foundational
7:05
Ejaaz: level and AI models. very much operate at that layer.
7:08
Josh: Yeah, I was excited to disagree with you, but I actually totally agree.
7:11
Josh: I trust Zuck's judgment, but again, he hasn't done anything like this before.
7:15
Josh: He's never rolled out actual infrastructure. He's never rolled out tools at the scale.
7:18
Josh: So I guess the answer is we will wait and see.
7:21
Josh: But that's not all the news for this week. What do we have next?
7:23
Ejaaz: Well, I kind of just wanted to kind of rain on your parade a little bit,
7:28
Ejaaz: Josh, because I know you're such an Apple fan.
7:30
Ejaaz: But despite, again, Meta bleeding so much, despite them losing so much money
7:35
Ejaaz: on this talent acquisition, they still have, you know, enough money to poach
7:42
Ejaaz: Apple's best in terms of AI.
7:45
Ejaaz: This broke this week that Zuck poached Apple's robotics chief from Meta's robotics
7:49
Ejaaz: studio. I just thought that in the midst of all of this happening,
7:53
Ejaaz: Apple still somehow bleeds, even though they are not at the frontier of AI yet.
7:58
Ejaaz: And especially because Tim Cook is probably the only other guy that could compete
8:03
Ejaaz: with Zuck at the consumer AI app level.
8:05
Ejaaz: So I don't know. I just thought it was pretty funny. Yeah.
8:08
Josh: The robotics chief implies that there is a robotics segment of Apple that I
8:13
Josh: think the whole world is just grossly unaware of.
8:14
Josh: So I guess we don't know how important this will be.
8:18
Josh: Maybe they're not even going to build. I mean, they were supposed to make a
8:20
Josh: car. They were going to make a TV. They had all these things in R&D.
8:23
Josh: Who knows how much that really means, but that's not the only company that
8:26
Josh: OpenAI has acquired, correct?
8:28
Ejaaz: No, it seems like there's been a bit of a snowball effect for the rest of the AI industry.
8:34
Ejaaz: Elon Musk, you know, owns and started XAI, which is another frontier AI model competitor.
8:40
Ejaaz: Previously been like the darling child on their show, actually.
8:43
Ejaaz: I think you and I will agree that we both love Musk. We love what he's doing with XAI.
8:47
Ejaaz: This startup is less than two years old, and they already caught up to the frontier.
8:51
Ejaaz: Grok4 is performing at, you know, crazy feats.
8:53
Ejaaz: But even they are leaking talent.
8:56
Ejaaz: One of the main stories is they're suing this guy called Shui Chen Li,
9:00
Ejaaz: I hope I pronounced that correctly, for allegedly stealing the entire Grok 4
9:06
Ejaaz: code base and then moving to OpenAI for a job.
9:11
Ejaaz: This quickly moved from being allegedly to being confirmed.
9:16
Ejaaz: Elon Musk actually confirmed this in a tweet. I was trying to find it, but I couldn't find it.
9:19
Ejaaz: That he has proved in the logs that he downloaded the entire code base and is
9:23
Ejaaz: moving to OpenAI. So obviously, this is a massive copyright infringement and
9:26
Ejaaz: concern for Musk and XAI at large.
9:30
Ejaaz: And so he's trying to sue him and prevent him from joining OpenAI.
9:34
Ejaaz: And just to add kind of like a little bit more spice to this whole affair,
9:38
Ejaaz: this employee in question also cashed out $7 million worth of XAI shares during
9:44
Ejaaz: the move. So just an insane turnaround of events.
9:47
Ejaaz: And again like just a reminder here like these guys are being traded around
9:51
Ejaaz: like some of the best super sports superstars ever right we're talking about
9:55
Ejaaz: 100 to 300 million dollar contracts and they're cashing out millions as they're
9:59
Ejaaz: doing so this might be actually the quickest way that you could make money in
10:03
Ejaaz: any kind of industry void of just tech alone um i just found this insane.
10:07
Josh: Yeah it's pretty amazing that uh he was able to cash out seven
10:10
Josh: million dollars invested stock after just one year of being at the company and
10:14
Josh: then move over to open ai but this this is an interesting idea that i've been
10:18
Josh: thinking about a lot which is um just the the fragility of these companies and
10:23
Josh: the ip that exists within these companies because a lot of the breakthroughs
10:26
Josh: that happen exist on a code base and
10:29
Josh: Perhaps the breakthrough is as simple as like 20 lines of code.
10:32
Josh: And it's just a small iterative thing that changes how a transformer works.
10:36
Josh: And those 20 lines of code that can optimize and save billions of dollars of
10:41
Josh: GPU training space, or it can optimize and create a model that blows all of
10:44
Josh: the others out of the water.
10:46
Josh: It's literally 20 lines of words. And if someone is able to get into a company,
10:51
Josh: figure out what those 20 lines are, extract them and make them public or give
10:55
Josh: them to another company, then you've essentially devalued billions of dollars
10:59
Josh: of value from a company by taking 20 lines.
11:01
Josh: And what we saw here is not just 20 lines, but the entire code base.
11:04
Josh: So I think this is to the detriment of a lot of companies because it's going
11:09
Josh: to be very challenging to battle on the frontier of actual AI efficacy versus
11:14
Josh: building applications and value for the users on top.
11:17
Josh: It's as fragile as a couple lines differentiating between the best models in
11:21
Josh: the world versus just the average company. And I I wonder what OpenAI is going to do with it.
11:24
Ejaaz: I was just thinking like, how valuable are your employees and how much kind
11:30
Ejaaz: of restriction can you place on them or should you place on them?
11:32
Ejaaz: I remember us doing an episode like four months ago, maybe five months ago,
11:37
Ejaaz: Josh, kind of shitting on Sam Altman's approach to his employees.
11:41
Ejaaz: We were like, he's being too constrictive.
11:43
Ejaaz: He's not allowing them to sell shares. These are the stars of the show.
11:46
Ejaaz: And he's placing this kind of like garden leaf ban on them. So if anyone leaves
11:52
Ejaaz: OpenAI, you can't basically join any major frontier lab for a couple of years.
11:56
Ejaaz: And I remember he eventually overturned it because he faced a lot of pressure in the media.
12:00
Ejaaz: We were kind of very bearish against Sam Altman and OpenAI. Turns out he might
12:04
Ejaaz: actually have been right.
12:05
Ejaaz: Because as you said, like, whether these guys download the code base or not, they have it in here.
12:09
Ejaaz: They know what kind of like parameters they're using, how they're training the
12:13
Ejaaz: model, how much compute they're using.
12:15
Ejaaz: And so these individual employees become the most valuable assets of these companies.
12:20
Ejaaz: And if you aren't placing restrictions on it, it becomes way too easy to poach
12:24
Ejaaz: and then suddenly become the frontier lab.
12:26
Ejaaz: And, you know, we've seen Meta poach a lot of OpenAI. We also know that X has
12:31
Ejaaz: poached a few open AI employees in the past, not to the extent that Meta has,
12:34
Ejaaz: and that Anthropic has done the same and back and forth and vice versa.
12:37
Ejaaz: So, you know, this becomes kind of like the main thing. And X isn't immune to this, right?
12:42
Ejaaz: You know, I've got up on our screen here and yet another employee announced
12:45
Ejaaz: that they were transitioning and moving on.
12:47
Ejaaz: So I'm just seeing this general trend of these AI researchers being the stars of the show.
12:52
Ejaaz: They know their value. I think some of them, and this might be a hot take,
12:56
Ejaaz: are just cashing out on money because they probably see some kind of AI bubble forming.
13:00
Ejaaz: But the other side of me also tells me that they kind of want to work on something
13:03
Ejaaz: that is meaningful and purposeful for them.
13:06
Ejaaz: We saw that with Zuck offering, I think, what was it, $3.5 billion to poach
13:12
Ejaaz: the CTO of Thinking Labs or Thinking Machines, rather, which is Mira Murati,
13:17
Ejaaz: the ex-co-founder of OpenAI, which is setting up her own AI thing.
13:22
Ejaaz: He rejected it, right? He was like, no, I kind of want to work on this vision
13:25
Ejaaz: and purpose. So I see those two conflicting views.
13:28
Ejaaz: But the rapidness of what I'm seeing here of employees kind of moving from company
13:32
Ejaaz: A to company B after two weeks just kind of sounds like a cash grab to me.
13:36
Ejaaz: But maybe that's just because of my experience in crypto. Maybe I'm too cynical.
13:40
Josh: Yeah, well, I mean, the money was flowing around not only with employees,
13:43
Josh: but also with other companies, too. OpenAI was in the news for another reason
13:47
Josh: this week, which is their acquisition of Statsig.
13:51
Josh: For people who don't know Statsig, I just learned this week because I didn't.
13:55
Josh: Statsig, and Ejaz, please correct me if I'm wrong, but I believe is kind of
13:57
Josh: like an A-B testing company, which led me down a lot of rabbit holes.
14:02
Josh: So basically what models can do that integrate Statsig is they're able to kind
14:05
Josh: of serve two results and then people can choose the better result and you kind
14:09
Josh: of iteratively get better through this system.
14:12
Josh: Now, I was thinking of like, why would they want this, right?
14:14
Josh: Why do they want to have these a b testing well one is just
14:17
Josh: to make the model better one is just to improve outputs but also i mean
14:20
Josh: benchmarks are getting a little stale right uh benchmarks don't work that well
14:24
Josh: people are kind of maxing out on them so perhaps they kind of create their own
14:27
Josh: new benchmarking system where they can compare not just their model results
14:31
Josh: to um themselves but also to other ais so i'm curious what do you think what
14:36
Josh: was the reasoning behind purchasing static it's a
14:38
Ejaaz: Huge amount of money and also they did that to pretty much acquire one person
14:42
Ejaaz: by the And we can dig into that, into Vijay, who is now the new CTO of OpenAI
14:46
Ejaaz: applications in a second.
14:47
Ejaaz: The reason I think they spent that amount of money is, well,
14:51
Ejaaz: number one, to your point, Josh, benchmarks are terrible. I don't care what anyone says.
14:55
Ejaaz: These AI benchmarks were maybe like good a year ago. Now they suck.
14:59
Ejaaz: Now every Frontier AI model lab is tweaking or fine tuning the new models that
15:04
Ejaaz: they release just so that it passes an exam.
15:06
Ejaaz: And I'm not really interested in that. I want to know how effective it is in real life.
15:11
Ejaaz: And actually to use the example of Meta, who we were just talking about,
15:15
Ejaaz: Void of AI, every new social media feature that they've launched on Instagram,
15:19
Ejaaz: on WhatsApp or whatever, Josh, they're A-B testing it across 10 different countries
15:23
Ejaaz: or communities every single moment.
15:25
Ejaaz: I've actually been privy to this, right? Where I've got a new feature on my
15:28
Ejaaz: phone and my girlfriend hasn't, right?
15:29
Ejaaz: So they're constantly testing things to see what works and what doesn't and
15:33
Ejaaz: how to like properly launch a feature.
15:35
Ejaaz: This is incredibly hard to do in a landscape or tech sector that's never been tested before, right?
15:42
Ejaaz: And AI moves so quickly. So how do you know what to tweak, when to tweak,
15:47
Ejaaz: and who to pitch it to are all questions that no one has any idea how to answer.
15:53
Ejaaz: And StatSig basically formalizes all of that, and they make a ton of money doing
15:57
Ejaaz: so. One thing that I found interesting is.
16:00
Ejaaz: $1.1 billion was used to buy this guy called Vijay, right? And you might be
16:05
Ejaaz: wondering, well, who the hell is Vijay?
16:07
Ejaaz: He is a tenured employee at Meta in his history before he started Statsig.
16:12
Ejaaz: In fact, he was doing this exact job that Statsig does at Meta for a decade.
16:17
Ejaaz: And he was working alongside a little someone called Fidji Simo.
16:22
Ejaaz: And if that name sounds familiar, that's because Fidji was appointed as the
16:26
Ejaaz: CEO of OpenAI Applications, of OpenAI applications, not OpenAI,
16:31
Ejaaz: that's still Sam Altman,
16:32
Ejaaz: but she's the head of basically all the new consumer applications and enterprise
16:36
Ejaaz: applications that OpenAI is going to launch over the next couple of months.
16:39
Ejaaz: So it's a pretty big role.
16:41
Ejaaz: And they worked alongside with this
16:43
Ejaaz: guy called Vijay, AB testing all the applications that she did at Meta.
16:46
Ejaaz: And so she was probably thinking, hey, Sam, we're kind of reaching the point
16:50
Ejaaz: where we're going to launch a bunch of really cool things at OpenAI.
16:53
Ejaaz: I need a guy like Vijay here to do this. And he was like, okay, name your price.
16:57
Ejaaz: Should we just acquire his company and his tooling? is that worth
17:00
Ejaaz: it and she probably said yes and that's pretty much what
17:03
Ejaaz: has happened um and then you might be thinking well
17:06
Ejaaz: okay jazz you've just said your thesis but
17:09
Ejaaz: is there any proof behind this like who were the customers of this
17:12
Ejaaz: well that's pretty big customers one of their biggest actually being anthropic
17:17
Ejaaz: um who i don't know how much they were paying them per year but um it was for
17:23
Ejaaz: this thing called telemetry services where basically clawed code which is anthropic's
17:27
Ejaaz: main and product on what's made them super famous.
17:30
Ejaaz: Is plugged into Statsig to A-B test a bunch of coding features that they're launching.
17:36
Ejaaz: And so you could think that Statsig itself has acquired quite a hugely valuable
17:41
Ejaaz: data reserve of all of Claude Code's users and behaviors.
17:45
Ejaaz: Now it says here in this screenshot that apparently the data is encrypted.
17:50
Ejaaz: I don't know how true that's gonna be. And you know what this reminds me of,
17:53
Ejaaz: Josh, just to kind of like finalize my point here,
17:56
Ejaaz: Do you remember when Meta invested $15 billion in scale AI? We just spoke about it.
18:01
Ejaaz: Do you remember who their biggest customers were? It was Google.
18:05
Josh: It was OpenAI. Yeah. It's also incestual.
18:09
Ejaaz: And they pulled their customers immediately. Yeah, it's crazy.
18:11
Josh: Yeah, it is crazy. What I understand is the data itself is encrypted,
18:15
Josh: but that encryption only goes down to a certain level and you still get a lot
18:21
Josh: of the top level data that is valuable.
18:22
Josh: It just lacks the personal data. so they still they do collect a lot of data
18:26
Josh: for this i love that open ai now is assigning
18:30
Josh: sub ceos to the ceo and building out actual executive
18:33
Josh: and c-suite teams around just creating applications and
18:36
Josh: products because i mean like we mentioned a little bit earlier these models
18:39
Josh: probably become commoditized at some point where like
18:43
Josh: they just everyone squeezes out the same efficiencies they're all pushing towards the
18:45
Josh: same frontier eventually the competition is going to happen on
18:48
Josh: the app layer and open ai is i mean they are probably ahead
18:51
Josh: in this by a long a large margin where they
18:54
Josh: actually have an entire c-suite dedicated specifically to
18:58
Josh: creating these applications creating a better user experience and
19:01
Josh: i think that's part of the reason why i mean the rest of the world they have
19:03
Josh: the most users but also why we keep coming back why i keep coming back is their
19:07
Josh: applications just they rock and if they're going to stack a hardware device
19:10
Josh: on top of that and they're going to keep rolling out these cool new apps i mean
19:13
Josh: sign me out i'm bullish on the ai open ai but again this is not the only crazy
19:17
Josh: news for this week we have another big announcement uh
19:20
Ejaaz: Yeah so anthropic casually raised 13.1 billion i think josh this week in a series
19:28
Ejaaz: f which values them at drumroll 200 billion dollars this is a private company
19:36
Ejaaz: what sorry hang on i'm gonna stop talking for a second like what are we doing here.
19:40
Josh: 183 billion dollars okay this
19:44
Josh: is fascinating so from from may of 2023 to
19:47
Josh: september of 2025 which is two years and change um they went from a four billion
19:52
Josh: dollar valuation to a 183 billion dollar valuation that is a multiple oh my
20:01
Josh: god what 20x in two years i can't
20:04
Ejaaz: Do math i can't yeah.
20:05
Josh: No 20 no 20x would be 80 that'd be
20:10
Josh: 40x, 50x?
20:11
Ejaaz: Tune in on Josh and I just being one-shotted by ChatGBT and we can't do basic mental math.
20:18
Josh: We're going to go with roughly 45x. That's my guess. In two years,
20:22
Josh: in 24 months, that is outrageous levels of growth.
20:25
Josh: And I mean, it leads everyone, I think, to ask themselves, how much further can this really go?
20:31
Josh: Like, where does the strain stop? Can they really do another?
20:35
Ejaaz: Okay, Josh, I'm just going to ask you flat out. Are we in a bubble?
20:38
Josh: I think we are. In some instances, and in some instances, we're not.
20:42
Josh: So this very much feels unsustainable.
20:48
Josh: We cannot, Anthropic will not grow, cannot keep doubling from here.
20:51
Josh: I mean, two more doubles and you're at a trillion dollars, basically.
20:55
Josh: But AI is the most formative tech of our generation by a pretty large margin.
21:02
Josh: And it's not going anywhere, and it's going to keep providing lots of value.
21:05
Josh: So where everything is valued
21:08
Josh: at now is probably very high for the
21:11
Josh: current time yeah but where it will be at relative
21:14
Josh: to now and even just five years from now feels
21:17
Josh: very undervalued so it's probably a matter
21:20
Josh: of of the rate at which we're accelerating which probably
21:23
Josh: is too high it probably is some sort of a bubble but i'm not sure the bubble
21:26
Josh: necessarily needs to burst it is just really this this pivotal transformative
21:31
Josh: thing that takes decades to play out and we are in we're still decade number
21:35
Josh: one so it doesn't it feels like there is there's overvaluation but the long-term
21:42
Josh: effects of this are just up only i think that's yeah that's my take at least do you have any
21:49
Ejaaz: I do have different dates. I agree with you on the long term,
21:52
Ejaaz: but I think we are in a pretty big bubble.
21:55
Ejaaz: I don't think we're anywhere near to it popping. I think things are going to get a lot crazier.
22:00
Ejaaz: I think we're going to reach trillions, like double digit trillions, dollar market caps.
22:06
Ejaaz: Nvidia is well on its way there. I think that's like a 2x, maybe less than a 2x from here.
22:11
Ejaaz: But I was also looking at the fundamentals of this anthropic raise, Josh.
22:16
Ejaaz: Um january this year they
22:19
Ejaaz: were making one billion dollars annual recurring
22:23
Ejaaz: revenue now simply eight months later they're now making five billion dollars
22:28
Ejaaz: of annual recurring revenue and i'm looking at this valuation and i use my trusty
22:33
Ejaaz: calculator that is a 36.5 x value multiple to their ARR, right?
22:40
Ejaaz: Which I don't know if that's a lot or if that's overvalued.
22:45
Ejaaz: I would guess it probably is. But like you said, AI is the darling chart.
22:51
Ejaaz: Literally, the MAG7, so the top seven companies in the S&P 500,
22:55
Ejaaz: are purely responsible for keeping the S&P 500 positive this year.
22:58
Ejaaz: And it just broke all-time highs.
23:00
Ejaaz: So on one side, you could argue, it's massively inflated.
23:03
Ejaaz: But on the other side you could say well if it wasn't for that the economy would
23:07
Ejaaz: be dying right now so you know it remains to be seen but i'm going to call bubble
23:11
Ejaaz: at this point i don't know when it's going to pop but but i'll take the bubble side all.
23:14
Josh: Right bubble it is so on the topic of bubbles and
23:17
Josh: sharks that go up into the right we have another one that is brought
23:21
Josh: to us today by grok our good old friends at
23:23
Josh: xai which just hit the number one
23:26
Josh: model on open router grok code this is
23:29
Josh: a brand new model that they just released and my understanding is
23:32
Josh: it is incredibly lightweight very fast and
23:36
Josh: very cheap and as a result what we're seeing here is
23:39
Josh: 96.5 billion tokens generated
23:42
Josh: by the model making it up 53 percent relative
23:46
Josh: to others and taking the number one slot on open router now for those who don't
23:49
Josh: know about open router we actually had the ceo on a few weeks ago and the way
23:53
Josh: open router works developers come they plug into the open router infrastructure
23:56
Josh: and then the actual router decides where to route traffic through based on the
24:01
Josh: needs of the developer. So what we're seeing here is
24:04
Josh: developers are requesting a model from Open Router and Open Router is deciding
24:08
Josh: over and over and over again that the best model to serve the users is Grok Code.
24:14
Josh: Now, Ijaz, do you have an idea why this is, what makes this so special,
24:17
Josh: why it's gone from nothing to a huge deal?
24:19
Ejaaz: I remember when Alex, the CEO and founder of Open Router, came onto our show
24:24
Ejaaz: and we asked him this exact question.
24:26
Ejaaz: We were like, you know, why do some of these models that get listed get used so aggressively?
24:29
Ejaaz: He said that he has the largest community
24:33
Ejaaz: of ai developer nerds that anyone's
24:37
Ejaaz: ever seen so if you want to test your product or
24:40
Ejaaz: your new model specifically the coding element you kind
24:43
Ejaaz: of want to be on open router you kind of want to access that community and so
24:46
Ejaaz: i think this is a positive signal be it
24:49
Ejaaz: small and kind of concentrated and i think we're actually going to see the kind
24:54
Ejaaz: of wider effects of how good this AI coding model is in the next couple of months
24:59
Ejaaz: when regular developers that are working in completely different or maybe even
25:03
Ejaaz: adjacent tech sectors start using it for their work that isn't AI specific, if that makes sense.
25:09
Ejaaz: And you mentioned, Josh, that, you know, it was 96 billion tokens.
25:13
Ejaaz: They actually crushed that record. I think it was yesterday or two days ago
25:17
Ejaaz: where they set a new all-time record of 138 billion tokens in a single day.
25:24
Ejaaz: Um that is insane that that metric
25:27
Ejaaz: has literally not been achieved with claude code's uh
25:30
Ejaaz: sonnet model which was previously the kind of darling child
25:33
Ejaaz: that wasn't even achieved on codex which is open
25:36
Ejaaz: ai's coding darling so um there's
25:39
Ejaaz: a lot of demand for this thing and i remember when they released croc 4 josh
25:43
Ejaaz: um croc 4 uh alone um its coding function was amazing and it beat claude sonnet
25:49
Ejaaz: so the fact that they have like released another coding model so soon so quickly
25:53
Ejaaz: after indicates that this is a niche within AI.
25:57
Ejaaz: That is progressing,
25:59
Ejaaz: potentially faster than the actual llms themselves and something to keep an
26:03
Ejaaz: eye on i i my bet is we're going to see a software coding agent actually outperform
26:09
Ejaaz: senior to mid-level engineer mid-level to senior engineers much sooner than we anticipate it's.
26:15
Josh: Amazing to see their name higher than claude sonnet
26:18
Josh: gemini 2.5 flash gemini 2.0 i mean all
26:21
Josh: the leading models again after only existing for two years
26:24
Josh: and change and i'm going to sound like a broken record but
26:27
Josh: xai's rate of acceleration is unbelievably fast
26:31
Josh: the rate that they're shipping these models but also the
26:34
Josh: rate that they're shipping models that matter i think
26:37
Josh: no one would argue that this is an actual better
26:40
Josh: model than claude sonnet 4 which is kind of like the
26:43
Josh: leading edge of code generation but they're optimizing for
26:45
Josh: things that developers want they're optimizing for cost for efficiency and
26:49
Josh: for just pretty good not great and i'm
26:52
Josh: sure the great will come and i'm sure claude will then work on
26:54
Josh: the pretty fast and pretty cheap but what they're doing
26:57
Josh: is they're optimizing for the things that clearly developers want the
27:00
Josh: most instead of going for the benchmark so i think while claude
27:03
Josh: and gemini we mentioned a lot of these models were optimizing for benchmarks this
27:06
Josh: is probably optimizing for raw utility and that's what
27:09
Josh: we're seeing coming out of these tokens uh being at
27:12
Josh: all-time 138 billion in a day that's outrageous
27:15
Josh: crazy crazy and then
27:18
Josh: for our last news of the day apple we have
27:21
Josh: something about apple and it's positive it is not something negative i promise
27:25
Josh: you it's actually cool it's exciting apple is doing something interesting in
27:28
Josh: the world of ai and maybe this is a teaser for next week next week they have
27:32
Josh: their iphone release event where they're announcing all the new hardware we
27:35
Josh: will be covering that but before that they have dropped two new models.
27:39
Josh: Ijaz, tell us what's going on here. Apple is now in the news for AI.
27:43
Ejaaz: Okay. I'm glad that I can say something positive about Apple in the AI realm
27:47
Ejaaz: because it has been excruciatingly difficult to do that in the past.
27:52
Ejaaz: They've done a few things here that puts a massive grin on my face, Josh.
27:57
Ejaaz: Number one, they've released two, not one, but two AI models.
28:01
Ejaaz: And they've open sourced both of them.
28:04
Ejaaz: You know, I'm a massive open source fan. I believe that This technology should
28:07
Ejaaz: be for everyone and everyone should be able to modify it to their own tailored
28:12
Ejaaz: niche and use case or whatever that might be.
28:14
Ejaaz: It's a giving good that a lot of these tech monopolies don't usually do.
28:18
Ejaaz: And I'm happy to see this kind of take place. But then the next question is,
28:22
Ejaaz: OK, well, what do these models actually do?
28:24
Ejaaz: Is it just another LLM? Is it going to be worse than ChatGPT?
28:27
Ejaaz: Well, they actually did something pretty novel here.
28:30
Ejaaz: They released two models, the first one being something called FASTVLM,
28:34
Ejaaz: which stands for vision learning model.
28:37
Ejaaz: Basically is an AI model that combines vision, which is like looking at images
28:42
Ejaaz: with the ability to understand text or communicate what it's seeing via text.
28:48
Ejaaz: So if you look at this demo here, you'll see it's a video of some kind of like Apple thing happening.
28:55
Ejaaz: There's like a race car happening. There's robots falling over.
28:58
Ejaaz: And the ability for an AI model to kind of like see what's on its screen and
29:03
Ejaaz: accurately depict what's going on and communicate that is actually a very underrated skill.
29:09
Ejaaz: The reason being is it's easy to kind of film a video and kind of like feed that to an AI model.
29:14
Ejaaz: Very, very hard to get that AI model to understand what the hell's going on in that video.
29:19
Ejaaz: That's why we've seen a slower rate of progression with AI video models or AI image models.
29:24
Ejaaz: It's taken a bit of time versus LLMs, which are getting kind of like to PhD
29:28
Ejaaz: level status here, right?
29:29
Ejaaz: But it seems like Apple has been kind of working behind the scenes on a model
29:33
Ejaaz: that does this in a super refined manner.
29:36
Ejaaz: And these words that it's generating, you can see it on the screen,
29:39
Ejaaz: if I can actually just blow it up, it's doing it in real time,
29:42
Ejaaz: which is incredibly hard to do.
29:43
Ejaaz: You know, it's saying, okay, there's a race car, Tim Cook has headphones on,
29:47
Ejaaz: and he's speaking into the mic, talking to his employees, that's kind of talking about this race car.
29:52
Ejaaz: And the point is these words can now be fed back into a model or an llm to make
29:57
Ejaaz: it super refined and much smarter.
29:59
Josh: So why is this important why does this matter because i
30:02
Josh: think this technology you just it's existed before like models
30:05
Josh: can do this models can see videos they understand this is not a this at this
30:09
Josh: point it's kind of a trivial thing for an ai model to do i think what's interesting
30:12
Josh: with apple and the reason why it's important and the reason why apple's model
30:16
Josh: is different is because of how efficient and lightweight the model is when you
30:20
Josh: think of apple they have their iphones they need models that can run on a mobile phone.
30:24
Josh: You cannot run any of these models that we've talked about today on anything
30:27
Josh: less than a hardware rack that would fit in like the size of an office,
30:31
Josh: like a gigantic computer.
30:32
Josh: These models, they can run on your phone. And it says in this post here,
30:36
Josh: the models are up to 85 times faster and 3.4 times smaller than previous work,
30:41
Josh: enabling real-time vision language models. So this is cool.
30:45
Josh: This is interesting because it's applicable to us. This will be a model that
30:49
Josh: can run on an iPhone. This can be a model that delivers on the promise of Apple
30:53
Josh: intelligence that can be very lightweight and you could use it in your day-to-day life.
30:56
Josh: So an example that I really like is that let's say you're at a restaurant and
31:00
Josh: you have this gigantic menu with tons of stuff on it and all you have is your
31:03
Josh: phone and maybe the menu isn't immediately accessible online.
31:06
Josh: Well, you could just point your phone at the menu.
31:08
Josh: It'll read all of the items and you could ask it, hey, what's the healthiest
31:13
Josh: item based on my current diet?
31:14
Josh: Or what's a gluten-free option for the person who doesn't have gluten at the
31:19
Josh: table? And it can kind of be this...
31:22
Josh: This enhanced system similar to what we imagine the glasses will be like,
31:26
Josh: but currently in the form of a phone.
31:28
Josh: So I think in terms of usability, this is much more practical than just about
31:32
Josh: any other model we've covered today.
31:34
Ejaaz: So I have an additional kind of thought about this, which is Apple probably
31:40
Ejaaz: is one of the largest, if not the largest, holders of photo and video data.
31:49
Ejaaz: ICloud, right? Everyone stores everything on iCloud. And I wonder if there is
31:53
Ejaaz: an angle here where Apple is like, hey, we've got all this data.
31:58
Ejaaz: I'm not just going to feed this to a Google or an open AI to train their own models,
32:04
Ejaaz: but we can train our own models or use that data to kind of sell and kind of
32:09
Ejaaz: like negotiate a specific partnership with Google to kind of use,
32:13
Ejaaz: we get to use the AI model,
32:15
Ejaaz: but exclusively using this data and this
32:17
Ejaaz: feature that means apple gets a stake in this i i'm
32:21
Ejaaz: kind of like um hypothesizing here because rumors have been going around that
32:25
Ejaaz: um they might be partnering with google's gemini ai model or open ai's chat
32:29
Ejaaz: gpt because i think they're kind of a bit behind to create their own foundational
32:33
Ejaaz: model so that's a kind of like moonshot idea but we'll see if it maybe plays out i don't know.
32:38
Josh: Yeah we're seeing this kind of split pronged approach where they'll they have
32:41
Josh: the local compute and then they will have the cloud compute and it's
32:44
Josh: becoming more and more clear that they'll probably have to offload that cloud compute the
32:47
Josh: high computation stuff that takes a lot of time to think that requires a
32:50
Josh: big model it's probably not going to be theirs they'll probably i
32:53
Josh: mean they've been deferring to open ai maybe they'll do google they'll outsource
32:57
Josh: that but the actual local models that run on the phone well they can get pretty
33:00
Josh: good and they could still manage to deliver a lot of the promises that they
33:04
Josh: they offered us at wwdc last year of apple intelligence using things like these
33:08
Josh: new open source models so it's it's very much a step in the right direction
33:11
Josh: it is not a seal of approval we
33:14
Josh: We are pro-Apple in terms of
33:16
Josh: their AI capabilities, but it is definitely a step in the right direction
33:20
Josh: that we like to see, right?
33:22
Josh: They're trying, they're doing something. I mean, the bar is literally on the floor.
33:26
Ejaaz: Well, it is Tech-temper, as you said, Josh. That it is.
33:30
Ejaaz: We have a huge month. This is traditionally the month where all the big dogs,
33:35
Ejaaz: all the big tech companies come out with their latest releases.
33:38
Ejaaz: This is across hardware and software.
33:41
Ejaaz: And I'm excited for a new category of hardware around robotics.
33:45
Ejaaz: We saw Meta is kind of like poaching robotics executives from Apple and stuff like that.
33:50
Ejaaz: So I think it's going to be a really exciting month. I'm hoping,
33:54
Ejaaz: no, I am praying that Apple comes out with a flagship AI model that crushes
34:00
Ejaaz: OpenAI, but I think I'm being a little too optimistic.
34:04
Ejaaz: I'm hoping Meta Superintelligence Labs stops bleeding their executive team and
34:10
Ejaaz: actually puts out something that I would wanna use along with some consumer applications.
34:14
Ejaaz: And I hope that a new burgeoning startup in the name of OpenAI and maybe even
34:19
Ejaaz: Anthropic that are so lowly valued at $200 billion to $500 billion,
34:24
Ejaaz: comes out with a suite of applications that can go head-to-head with some of
34:28
Ejaaz: the GOATs, Meta and Apple. That remains to be seen.
34:32
Ejaaz: Josh, I have nothing else to say.
34:34
Josh: Do you? I just want to fill people in on our schedule for this month because
34:38
Josh: it's packed. Next week, we have Apple.
34:41
Josh: Apple, big event. This is the iPhone event. This is where they're releasing a lot of new hardware.
34:45
Josh: The hope is that not only will we get hardware, but we will get supplemental
34:48
Josh: AI software. So a lot remains to be seen.
34:51
Josh: A lot will be unveiled next week. The following week, we have another exciting
34:54
Josh: announcement, which is through Meta Connect,
34:57
Josh: I believe is what the event is called, where Meta is going to release their
35:00
Josh: hardware products and most certainly will be paired with a lot of software and
35:04
Josh: AI offerings. So we will have all of the coverage of that.
35:06
Josh: And then a few weeks later after that, we have OpenAI's Dev Day,
35:10
Josh: which I'm sure they will announce even more features.
35:13
Josh: And then a couple weeks after that, we have Microsoft Ignite,
35:16
Josh: which is Microsoft's new hardware offering day.
35:18
Josh: So it is going to be like a very crazy next
35:21
Josh: couple of weeks and months uh so buckle up this
35:24
Josh: was the news up to today for this
35:27
Josh: week there's going to be a lot more we will be back for more
35:30
Josh: episodes later this week next week the following week it is there's a lot to
35:34
Josh: look forward to in this space and we will be right here along your side going
35:38
Josh: through it all the highs and lows the ups the downs um as always if you enjoyed
35:42
Josh: the episode please share it with your friends don't forget to like subscribe
35:46
Josh: all the good things and we'll
35:48
Music: Music