Latest / In the Loop AI / Zuck's Open Source Playbook: Meta Muse Glimmer and 14-Page Manifesto

Transcript

Zuck's Open Source Playbook: Meta Muse Glimmer and 14-Page Manifesto · In the Loop AI

0:00
Josh: Meta has pivoted again for the third time and now back to an open source company.
0:05
Josh: The story of Meta has been pretty insane.
0:07
Josh: Two years ago they were totally open source, hellbent on wreaking havoc on the
0:12
Josh: entire marketplace of closed source models.
0:14
Josh: They were doing this through their model called Llama and for those who weren't
0:16
Josh: familiar, Llama was just kind of like this close to frontier open source model
0:21
Josh: that was expected to change the world in a meaningful way.
0:23
Josh: It did not go as Zuck foresaw and therefore they went and they closed off all
0:28
Josh: their model development and went closed source just like all of the other labs now
0:32
Josh: they're going back they're transitioning back to open source and as a result
0:35
Josh: they've spent a tremendous amount of money on this so what happened in the last
0:39
Josh: earnings report after all the capex spend came in
0:41
Josh: zuck just dropped a 14 page paper explaining exactly why they're making a re-entry
0:46
Josh: into the open source world
0:48
Josh: That entry is defined by a singular model that we're going to talk about right
0:51
Josh: now called Muse Glimmer. And Muse Glimmer is a really impressive model that
0:55
Josh: fits on your laptop, runs locally, and I think it's going to shift the world of open source AI.
1:01
Ejaaz: I think this is a great move from Zuck and Meta. And this is coming from probably
1:06
Ejaaz: the Meta's biggest bear in the past on this show.
1:10
Josh: Yeah, definitely. We're professional haters. I should preface this with.
1:13
Ejaaz: I'm a professional hair of Zuck and Meta for the last six months,
1:17
Ejaaz: but I love to see this pivot.
1:18
Ejaaz: And I'll explain why. Number one, Metz's entire philosophy around AI has always
1:24
Ejaaz: been open source. And his simple reasoning behind that is he believes like everyone
1:28
Ejaaz: should get access to this thing, right?
1:30
Ejaaz: If everyone has access to this thing, there'll be no centralized core power
1:34
Ejaaz: for AI. There's no AI overlord going forwards. He then kind of pivoted on that,
1:37
Ejaaz: like you mentioned, and now he's back with not one, but two new open source
1:41
Ejaaz: models. So what are the models?
1:43
Ejaaz: Number one, the headline is called Muse Glimmer. It's an open-weight,
1:48
Ejaaz: open-source, 30 billion parameter model.
1:51
Ejaaz: Now, if you're not familiar with the sizes of these relative models,
1:54
Ejaaz: this is a pretty tiny model.
1:56
Ejaaz: It can actually fit and run on your MacBook. It's 20 gigabytes of data or RAM
2:01
Ejaaz: that it requires, which is significantly smaller than any other frontier model,
2:04
Ejaaz: which is like hundreds and hundreds and hundreds of gigabytes.
2:07
Ejaaz: And the way that they were able to achieve this is they took a bigger model,
2:11
Ejaaz: and they made it smaller. It's something called quantization.
2:14
Ejaaz: Now, why did they create this model?
2:16
Ejaaz: Well, he believes or Meta believes that locally run private AI models.
2:22
Ejaaz: Should be the future of how AI models are dispersed amongst everyone.
2:25
Ejaaz: Right now, we pay subscriptions to get access to a model that lives in the cloud,
2:30
Ejaaz: that lives in a centralized company like Anthropic, like OpenAI.
2:33
Ejaaz: With this model, you can kind of run it locally.
2:36
Ejaaz: And that brings several different advantages that we've mentioned quite a few
2:39
Ejaaz: times on this show. The first one being it's cheap as hell.
2:42
Ejaaz: The second one being it can run on your private data, so it can become a lot
2:45
Ejaaz: more smarter and attuned to you specifically.
2:49
Ejaaz: And the third thing is it's very efficient when it serves you answers and results
2:54
Ejaaz: to your different prompts. So I think it's a good model, but it's certainly not frontier in any way.
2:59
Josh: Yeah, it's far from frontier. And I imagine they're probably not really concerned
3:02
Josh: about that. We mentioned this on previous episodes where there is this fight
3:06
Josh: for the frontier, but now there are different frontiers.
3:09
Josh: There is the frontier of intelligence, which Anthropic and OpenAI are kind of
3:12
Josh: working on. Then there's the frontier of cost and efficiency per token.
3:16
Josh: That's kind of where I see meta of place in themselves. In fact,
3:19
Josh: they're fighting on multiple fronts at once.
3:21
Josh: And it seems like Muse Glimmer is sitting at the very base of that because it
3:25
Josh: is such a small model, it's so quantized, it's able to actually run on a local machine.
3:29
Josh: I think the fully capable largest version of this takes about 55 gigabytes of
3:34
Josh: memory, which runs on my MacBook that I have sitting on my desk right now.
3:37
Josh: And that's really impressive.
3:38
Josh: This puts them in a race that not many other companies are competing on,
3:41
Josh: which is just like that edge inference local compute thing that I think we can
3:46
Josh: expect to see out of Apple Fairleague soon.
3:48
Josh: Like with the new Siri going to be running on local models, this is very much
3:52
Josh: feels like a response to that.
3:53
Josh: Where now for the first time you're able to run this free inference that's fairly
3:57
Josh: capable it has a somewhat large context window i believe a couple hundred thousand
4:02
Josh: tokens i'm not sure the exact number but a couple hundred thousand
4:04
Josh: Tokens of context it has you know pretty good outputs in terms of tokens per second
4:09
Josh: and it runs at a level that you would expect a frontier model would have ran at
4:13
Josh: maybe say 18 months ago so it's not going to do any crazy difficult complicated
4:17
Josh: tasks but for something that can run on your machine you could use anywhere
4:21
Josh: in the world at any given time and you can trust it to manage all of your data.
4:25
Josh: So for example, if you have a machine runs a lot of sensitive information,
4:28
Josh: you want to put health information through that, you might not want to give
4:32
Josh: to a public facing model.
4:33
Josh: This is a really easy way of doing that in a way that we haven't really had before.
4:37
Josh: And I think that's kind of step one in this new strategy that Zuckerberg's going
4:41
Josh: for in, I guess you could call it like scorched earth 2.0, where why do you
4:45
Josh: open source models? Well, you want everyone to go off, build on them to remove
4:49
Josh: margins away from these frontier labs.
4:51
Josh: And this seems like the first version of that that people are are kind of
4:54
Ejaaz: Excited about and you see him apply this logic to quite a few of his models
4:59
Ejaaz: now like he also open weighted or he plans to open weight.
5:02
Ejaaz: Muse spark 1.2 which if you have been following along with the show
5:07
Ejaaz: was a brand new model metas frontier model that they released i think it was
5:11
Ejaaz: like a week and a half ago and what made muse spark 1.2 really special is it's a really good
5:17
Ejaaz: coding model in the way that they've kind of engineered AI agents to kind of
5:20
Ejaaz: like replicate amongst themselves,
5:22
Ejaaz: take your code and go work in parallel.
5:25
Ejaaz: So it's a really efficient way of working. And that's kind of like Meta's core
5:29
Ejaaz: thesis, I feel, with a lot of these models that they're releasing.
5:31
Ejaaz: It's like they're not trying to get the most intelligent model,
5:34
Ejaaz: but they're trying to produce the most cheapest and affordable model that still
5:38
Ejaaz: gets about 90 to 95% of the work done.
5:41
Ejaaz: And Zuck has said in many times over the last couple of months that this is
5:44
Ejaaz: Meta's core philosophy. They believe that locally private-run,
5:47
Ejaaz: effective AI models will eventually outcompete the frontier. Now.
5:51
Ejaaz: When you kind of compare these models, right, you're probably listening to this
5:54
Ejaaz: and you're thinking, okay, who cares? Meta has a model. I haven't used Meta
5:56
Ejaaz: AI intelligence in God knows how long.
5:58
Ejaaz: Why should I use this model? Well, if you are using, say, the Anthropic or OpenAI
6:03
Ejaaz: API and you're spending a couple thousand bucks a month, even a couple hundred
6:06
Ejaaz: bucks a month, if you're an enterprise that's spending 10 to hundreds of millions of dollars per year,
6:11
Ejaaz: this might be a model that might be more bang for your buck.
6:13
Ejaaz: And listen, you're not going to use it for everything. You're not going to get
6:15
Ejaaz: it to plan your entire company strategy.
6:17
Ejaaz: You're not going to get it to build a brand new app from scratch and expect
6:21
Ejaaz: it to be the best thing, but you'll use it as a workhorse, maybe in conjunction
6:25
Ejaaz: with some of these more intelligent models.
6:27
Ejaaz: So it's great to see Zuck kind of like reprise Meta's role in open source.
6:32
Ejaaz: If we remember, like in the past, it was Lama and then he kind of like pivoted
6:35
Ejaaz: and everyone was like kind of super angry about this. I'm glad that he's doing this.
6:39
Ejaaz: The question is like, why is he doing this, right? Like, you know,
6:43
Ejaaz: why would this make sense for Meta, which is a for-profit publicly traded company,
6:46
Ejaaz: he sucks, still wants to make a lot of money.
6:48
Ejaaz: How does this make any sense at all? And he kind of dropped his entire thesis
6:52
Ejaaz: yesterday when he announced that he's open weighting a bunch of these models.
6:56
Ejaaz: He goes, I believe everyone should have access to super intelligence.
7:00
Ejaaz: And I wrote a long piece about Meta's philosophy and values for building that world.
7:05
Ejaaz: And essentially like this future that he describes balances on like three core
7:09
Ejaaz: pillars. Number one is he believes everyone should be individually empowered,
7:13
Ejaaz: aka get access to their own private AI model and that it shouldn't be centered
7:17
Ejaaz: in specific frontier AI labs.
7:19
Ejaaz: Number two, he believes the thing that pushes humanity going forwards is invention.
7:23
Ejaaz: So humans' ability and intention to invent brand new things should drive who
7:28
Ejaaz: gets access to these different models. And that could be anyone and everyone.
7:31
Ejaaz: And the third thing is he just believes in a core balance of power,
7:34
Ejaaz: which is something ironically that adversaries over in China have been advocating
7:39
Ejaaz: for a while now. So it's nice to see like a frontier American AI lab advocate for the same.
7:44
Josh: It is, but it feels a little bit like I can't, I can't feel like this is a high
7:49
Josh: conviction post because the sentiment has changed so much.
7:52
Josh: It's not like he's like they started this way, but then they changed this way
7:55
Josh: when it wasn't working. And now that things are working, they're going to go
7:58
Josh: back and start to believe this. It feels like they're just not really standing strong.
8:01
Josh: On morals over time and this this could be the wrong guess but i mean he's he's
8:06
Josh: here saying basically the exact arguments that everyone's been arguing for a
8:09
Josh: very long time it's like we don't trust a single monoculture to handle this
8:12
Josh: we don't believe that the future of humanity is going to be
8:16
Josh: best served by two to three companies we we don't believe that any company who
8:20
Josh: thinks that ai will eliminate most jobs would rush to build it faster than anyone
8:25
Josh: else and these are the arguments that we've heard for a really long time and
8:28
Josh: we just get these echoed over and over again
8:31
Josh: as the company starts to fit more of this mold i mean the essay was interesting
8:35
Josh: what i will be most interested in is just seeing what the actual downstream effects are of
8:40
Josh: these new essays and like hopefully it seems like zuck is getting more confident
8:44
Josh: that they're going to figure this out that they're going to actually be releasing
8:46
Josh: models that can meaningfully
8:48
Josh: shift the way people engage with ai and this is probably like the starting gun
8:53
Josh: he's like hey new era new company new ethos we're going to start running with
8:57
Josh: this and see how it goes and so far it looks like they're kind of on their way
9:01
Josh: like things are pretty good i was looking at
9:03
Josh: Muse glimmer and i actually i want to try this one so my plan is to install
9:06
Josh: this on my local machine today and give it a go
9:08
Josh: because one of the things that is unique about this one is for people who are
9:11
Josh: still using open claw remember that thing um this actually is a very agentic
9:15
Josh: model that runs end-to-end for task completion
9:18
Josh: it uses agents it has tool use it has multi-step reasoning it has multimodal
9:23
Josh: input and output it seems like a really strong and powerful model yeah we could
9:27
Josh: see here kind of how good it is at creating visuals, how good it is in benchmarks.
9:30
Josh: And it's an actually usable model in a way that I don't think a lot of these
9:34
Josh: others have been. So I'm taking this 14 page paper with a grain of salt. I...
9:38
Josh: Read some of it. I didn't get my way through all of it. EJS,
9:41
Josh: I know you said you read the entire thing.
9:43
Josh: I'm excited for Meta. I want them to win. So I'm hoping this is the new starting
9:47
Josh: point for them to release better models, to really get back in the game,
9:51
Josh: to compete. Because, I mean, yeah, duopolies are less than ideal.
9:54
Josh: We want a lot of people competing for a lot of different parts of the pie.
9:57
Josh: And if Meta is going to come out here and do it and do so open source,
10:00
Josh: that's great. And you got to imagine Jensen sitting over here pretty happy too, Mr.
10:04
Josh: Open Source King, that now there's another company in the running for making
10:08
Josh: really compelling open source models in the United States.
10:10
Ejaaz: I think it's actions over words, though. Zuck has talked this talk for a while, you're right.
10:14
Ejaaz: But he's also the guy that's releasing like the frontier open source models in America.
10:19
Ejaaz: You see all the other AI labs. Look, I'll give you an example here.
10:22
Ejaaz: Satya Nadella, bunch of people signed that they're in favor of open source and
10:26
Ejaaz: open weight movements in general, but they haven't released any frontier open models themselves. So
10:31
Ejaaz: I look at actions over words and Zuck, yeah, he's flailed around quite a bit.
10:35
Ejaaz: And I've been a big critique of that, but he's come back.
10:39
Ejaaz: And I'm going to judge him by his actions going forwards and we'll see what
10:42
Ejaaz: he actually uses this for.
10:43
Ejaaz: Now, the bad case for this is this is like an old school business move that
10:49
Ejaaz: he's making, which is let's flood the market with intelligence that is basically free.
10:54
Ejaaz: So we commoditize the paid subscriptions that other frontier labs are paying for.
11:00
Ejaaz: And that brings down their value, that brings down their profit margins,
11:04
Ejaaz: their revenue, and it allows Meta more time to catch up. Now,
11:08
Ejaaz: is this a costly and risky endeavor? Yes.
11:11
Ejaaz: Meta has burned, what is it, like $30 billion in acquiring 150 staff to produce
11:17
Ejaaz: a model that is sub-frontier. If you were a shareholder, you're probably looking
11:21
Ejaaz: at this and thinking, Zuck, what are you doing?
11:23
Ejaaz: But if you look at what he's been investing in outside of that.
11:27
Ejaaz: I think he's on track to spend, I think, $210 billion this year through the
11:33
Ejaaz: first half of next year on Datacenter CapEx.
11:37
Ejaaz: He is one of the largest holders of NVIDIA GPUs. He's secured like one of the
11:42
Ejaaz: largest arsenals of NVIDIA Rubins, which is their next-gen GPU architecture.
11:48
Ejaaz: Second only to Elon Musk, you have to question, okay, what's he going to use
11:52
Ejaaz: all of this compute for? Well, he could rent it out, which is what SpaceX is
11:55
Ejaaz: doing and make a bag from that.
11:57
Ejaaz: But he's also probably going to train a bunch of frontier models.
11:59
Ejaaz: And the biggest signal for whether he's actually going to be able to pull this
12:03
Ejaaz: off is model release cadence.
12:05
Ejaaz: And in the last, I think 90 days, he's released two models, but he's going to
12:11
Ejaaz: release another two more over the next, I think it's like one to two months.
12:15
Ejaaz: I think you mentioned in some tweet thread, Elon Musk is planning to do the same.
12:18
Ejaaz: So I'm seeing this trend where the companies that, even though they're a little
12:22
Ejaaz: bit behind in frontier intelligence, who have the most compute,
12:25
Ejaaz: who have the most capital, are still able to kind of ship models iteratively
12:29
Ejaaz: and maybe respond to the market quicker than some of these other frontier labs. Now,
12:32
Ejaaz: I appreciate that that's optimistic thinking, but I do think Zuck is back in
12:36
Ejaaz: the game and I'm excited to see what these models are like. I want to play around
12:40
Ejaaz: with this Muse Spark 1.2 and this new one that I can run on my laptop and see
12:45
Ejaaz: what fun things we can come up with.
12:47
Josh: Yeah, I was trying to unpack the reasoning. I'm like, okay, why did he make
12:50
Josh: this pivot again? It seems like there's so much uncertainty in the strategy of meta so far.
12:55
Josh: And there really does appear to be some pretty compelling reasons that are kind
12:59
Josh: of underlying this decision.
13:01
Josh: Because when you think about meta the company i was like okay what does meta
13:03
Josh: actually do meta is just an attention salesman
13:06
Josh: their biggest part of the business is advertising and sales through that
13:09
Josh: all their services are mostly free to use you can pay for a blue check mark if you want
13:13
Josh: but most their users are free and they have this tremendous user base so they
13:17
Josh: can sell a subscription to those people but those people are not used to and
13:22
Josh: not very probably likely excited to pay
13:24
Josh: twenty dollars a month to use an ai system that can be offered for free because
13:28
Josh: you have to assume the average user doesn't need frontier class intelligence
13:31
Josh: so if the model is attention
13:34
Josh: And if you can improve your core business through AI, then that's probably what
13:38
Josh: they want to do. And that's probably why open source makes sense.
13:41
Josh: I mean, yeah, meta already announced it wants to use your AI chats to personalize the ads
13:46
Josh: and it wants you to use the AI to build a large profile on you, to build
13:50
Josh: a josh.md, an ejaz.md that has all of our preferences, all the things we care
13:55
Josh: about, that gets updated in real time, and that has an increasing amount of
13:59
Josh: ways in which you can collect context on you.
14:02
Josh: And if you are trying to do this then you really just want these models to be
14:05
Josh: free and accessible and open to collect as much data as possible to feel like
14:08
Josh: there is no friction to getting in the game and then also just to kind of like just chip away at the
14:14
Josh: the frontier labs margins where if there's someone who is deciding between a
14:17
Josh: 20 model or just going into their instagram app and using this cool new chatbot
14:22
Josh: that they have like chances are a lot of people are going to choose to use the
14:25
Josh: chatbot and that puts them
14:26
Josh: in a way i guess it's kind of unique to them where they're they're competing
14:30
Josh: on edge inference in a way somewhat with apple and what they will be doing with
14:33
Josh: the new siri but they're also competing with google in the ads business if
14:37
Josh: they're going to have just as much if not more data than google has
14:42
Josh: When it comes to search queries and like think about how custom the advertising
14:45
Josh: can be and not even advertising but content creation in general if meta understands
14:50
Josh: what type of content you like how you engage with certain types of content it
14:53
Josh: can serve this very custom profile that allows custom generated content just
14:57
Josh: for you all in the effort of selling you more ads.
15:00
Josh: And I think this is the reality of Meta is as they've tried to pivot to hardware,
15:03
Josh: they've done all these other exploratory things.
15:05
Josh: It still just comes down to the core business is selling people ads.
15:09
Josh: And it's like understanding who their user is, pairing them with a seller and
15:12
Josh: collecting commission off of those sales.
15:15
Josh: And I think this is probably just a natural extension of that.
15:17
Josh: And they're going to continue to make these products more compelling and more
15:22
Josh: capable through these very lightweight AI models.
15:24
Ejaaz: At the end of the day, Meta is obviously a for-profit company.
15:29
Ejaaz: They're trying to make money and ads, as you said, is like their main business.
15:32
Ejaaz: So, you know, they're definitely going to focus on that.
15:35
Ejaaz: But I also think in the wider market of AI, a bunch of things have shifted recently
15:40
Ejaaz: that we just can't ignore.
15:41
Ejaaz: The first one being, you don't need the most intelligent model to do anything
15:46
Ejaaz: and everything for you. I was catching up with a friend who is pretty high up
15:49
Ejaaz: at a top, I would say, Fortune 100 company.
15:53
Ejaaz: And I asked him, like, you know, how are you using AI these days?
15:56
Ejaaz: Like, do you use private models? Do you use the standard models from Anthropic and OpenAI?
16:02
Ejaaz: And he said, I've shifted completely to some of these open source models,
16:06
Ejaaz: as well as just cheaper models in general.
16:08
Ejaaz: And I get it to do the bulk of my work whilst I use Accord or GPT 5.6 to brainstorm
16:14
Ejaaz: kind of like the architecture of what I want to build.
16:16
Ejaaz: And I'm seeing this pattern reflected across a number of different enterprises.
16:20
Ejaaz: It's certainly been reported across social media like every single day.
16:23
Ejaaz: We're seeing the shift of companies thinking, I'm not going to just needlessly
16:27
Ejaaz: spend a bunch of money on frontier intelligence. I certainly will for some things,
16:31
Ejaaz: but I want to explore some of the cheaper options, right? And you can look at
16:34
Ejaaz: the business models of the frontier labs themselves, right?
16:37
Ejaaz: I think over the last 90 days, eight models have been released,
16:41
Ejaaz: three of them alone from OpenAI, right? Why didn't they just release GPT-5.6
16:46
Ejaaz: Sol? Why did they release Luna? Why did they release Terra?
16:49
Ejaaz: Why did they slash the price of Luna by 80% a month ago? Why did Anthropic yesterday,
16:56
Ejaaz: I think, announce that they're going to keep Claude Sonnet, I believe,
17:01
Ejaaz: at the cheapest price ever indefinitely.
17:03
Ejaaz: I think these companies are realizing that cheaper models are definitely applicable
17:07
Ejaaz: and have an advantageous incentive for the customers that they're trying to
17:11
Ejaaz: sell to, whether it's enterprises or consumer customers.
17:14
Ejaaz: And you need more bang for your buck. Zuck realizes this, and he's seizing the
17:19
Ejaaz: opportunity, but he's being more aggressive with the open source side of things.
17:22
Ejaaz: And I think Zuck has enough capital at Meta to be able to keep this train going
17:26
Ejaaz: for a while now. So I'm interested to see how what meta and specifically SpaceX
17:32
Ejaaz: looks like over the next couple of months.
17:34
Ejaaz: I think SpaceX is also teasing their next model and they've released like two
17:38
Ejaaz: over the last like three weeks today, which should hopefully be releasing.
17:41
Ejaaz: But yeah, I think there's something that
17:44
Ejaaz: they're going to pull off something cool.
17:45
Josh: Yeah i'm looking forward to it i'm looking forward to all the agents that they're
17:49
Josh: going to spawn from these new models because if you are working with agents
17:51
Josh: you know you know who we got to talk about and that's ledger our sponsor of
17:55
Josh: this episode um because they have this infrastructure for agents in which it's kind of three parts
18:00
Josh: agents propose then the human approves and then
18:03
Josh: the ledger signer enforces the ledger signer is this mostly comes in the form
18:07
Josh: of a hardware wallet in which you can approve or deny we talk a lot about security
18:11
Josh: and how a lot of these agents can kind of go rogue and do things you may not want
18:14
Josh: ledger offers an agent stack which is a series of open source tools that allows you to
18:19
Josh: evaluate what these agents are doing across the way you could tell it to do
18:23
Josh: things like rebalance your wallet
18:24
Josh: or adjust whatever parameters you want to adjust and then it will propose you
18:29
Josh: approve and the ledger signer enforces it works with cloud code
18:31
Josh: codex all of the places that you work with ai on a daily basis so thank you
18:36
Josh: so much to ledger for sponsoring this part of the episode you can get their
18:39
Josh: products at the link in the description below.
18:41
Josh: EJS, you mentioned capex spend, or you actually didn't mention capex spend.
18:45
Josh: You mentioned strong earnings, but I'm going to mention capex spend,
18:48
Josh: which is their free cashflow.
18:51
Josh: Like we have to talk about the earnings quickly because like we do want to get
18:54
Josh: a little bit economic here. Like how viable is this strategy?
18:56
Josh: Their free cashflow this year, which means basically the money that comes in
18:59
Josh: and out is down 91% year over year.
19:04
Josh: They're spending, They're spending big on a few key pillars.
19:08
Josh: One of them is this open source bet, which does not have an immediate ROI because,
19:12
Josh: I mean, again, they're offering these weights for free.
19:14
Josh: The other is on their hardware platform. One of the things that is like still
19:19
Josh: shocking to me, and I'm not sure that many people know about this,
19:22
Josh: Meta owns 70 to 80% of the entire AI glasses market.
19:27
Josh: And this is not a big market, but this is a market that is meaningful and that
19:31
Josh: probably shouldn't be overlooked.
19:34
Josh: This is done exclusively with a company named Luxottica. Together,
19:37
Josh: they have more than tripled Meta's AI glasses sales just in the year of 2025.
19:42
Josh: We haven't gotten 2026 numbers yet, but like that is kind of noteworthy it's interesting to see the
19:48
Josh: floor of the capex guidance being raised too in 2026 meta is guiding for 145
19:56
Josh: billion dollars of spend which is just this like unfathomably large number
20:01
Josh: and when you think about the places that it's going to land well it's certainly
20:04
Josh: open source it's certainly hardware but like what else are they really just
20:08
Josh: going to go all in on building out data centers and gpus to make larger models is that
20:13
Josh: is that the end game? Because I don't see them competing on the frontier.
20:17
Josh: So do they really need that much compute in order to compete for open source, open weight models?
20:22
Ejaaz: I think so. I think the number one factor that determines how good your AI offering is, is compute.
20:29
Ejaaz: At the end of the day, Elon's going to prove that with SpaceX.
20:31
Ejaaz: And I think Zuck is planning to do the same thing.
20:34
Ejaaz: We saw NVIDIA or Jensen Huang announced yesterday that he secured an additional
20:39
Ejaaz: $500 billion of revenue and capital investment for future compute.
20:44
Ejaaz: We see OpenAI partnering with NVIDIA for a similar amount, half a trillion dollars.
20:49
Ejaaz: You see Anthropic securing deals with Riot yesterday with an old cryptographic
20:53
Ejaaz: mining company to secure even more compute.
20:55
Ejaaz: Compute is the ultimate indicator on whether you can create a better model.
20:58
Ejaaz: It is the one thing that is consistent across all competitors in the AI race.
21:02
Ejaaz: And I think Zuck believes if he amasses enough of it, if he secures enough allocation
21:06
Ejaaz: of frontier NVIDIA GPUs, he'll be able to pull off and build better models.
21:12
Ejaaz: Now, whether the open source strategy plays out, again, it's that strategy that
21:15
Ejaaz: I think I mentioned earlier, which is flood the market with commoditized intelligence,
21:20
Ejaaz: cut down or cut into revenue and margins of your competitors.
21:24
Ejaaz: Yeah, Meta's CapEx takes a beating in the short term, but hopefully that plays
21:29
Ejaaz: better in the long term when you have personalized agents that all run on Meta's
21:32
Ejaaz: infrastructure and consumer apps.
21:34
Ejaaz: Also, remember, Meta has like 4 billion consumers right now.
21:37
Ejaaz: I was scrolling on Reels yesterday, which I don't like to admit,
21:41
Ejaaz: but a ton of them are now Meta glasses.
21:44
Ejaaz: I don't know if you've seen this, Josh, but it's this POV from a bunch of different people.
21:48
Ejaaz: And I think that they don't have the best hardware, but they have one of the
21:51
Ejaaz: most amazing distributions ever. I think 4 billion, and then you compare that
21:55
Ejaaz: to Apple, which has, I think, 3.5 billion live Apple devices right now.
21:59
Ejaaz: It's probably a lot more. I think that data is outdated.
22:01
Ejaaz: They have a real competing moat here. And I look at Google, who has recently
22:06
Ejaaz: gone into negative cash flow. Meta isn't quite there yet.
22:09
Ejaaz: So you could argue that they're being somewhat conservative.
22:12
Ejaaz: They have other businesses that are earning a ton of money that can support this for a while, but,
22:16
Ejaaz: They don't have infinite runway. And, you know, maybe we'll see Meta raising
22:19
Ejaaz: external capital similar to how Intel and Google has over the last couple of
22:23
Ejaaz: months. I don't know. We'll see.
22:25
Josh: Yeah. And I was going to ask the question whether or not this CapEx spend is
22:28
Josh: existential. Like what happens if they fail again?
22:31
Josh: Because this isn't their first time trying to get a frontier model that works
22:34
Josh: really well and the first time did not go as planned. And then I was thinking, well, you know what?
22:39
Josh: At least they have the GPUs and they could just go resell those and make a great business out of it.
22:42
Josh: So there's this really weird part of the market now where it's very advantageous
22:46
Josh: to just spend money on building data centers. And even if you can't figure out
22:49
Josh: how to use them yourself, you just sell them off to the next best guy and you
22:52
Josh: can make a tremendous amount of money from that infrastructure build out.
22:55
Josh: So I think at the very least, you could find a lot of solace in that idea in
22:58
Josh: terms of users. Like that's a great point that, I mean, 4 billion users is half of the planet.
23:02
Josh: There's not many companies in the world that I'm not sure if there's any companies
23:06
Josh: in the world with that quite as many users as meta, or there's a very small handful of them.
23:10
Josh: So if they're able to figure out how to, I mean, weaponize is like a terrible
23:14
Josh: word for this, but like able to leverage their AI tools to further create compelling
23:20
Josh: offerings on their social media platforms, like that's a pretty big value unlock.
23:23
Josh: So I'm excited. I'm cautiously optimistic on meta.
23:27
Josh: Like I'm still looking for signals that they are going to figure it out.
23:30
Josh: But like you mentioned, when you're watching reels, a lot of it are coming from
23:33
Josh: the glasses. It's this novel form of creating content. It's this cool new paradigm
23:36
Josh: that every other company is working for. And right now, I mean,
23:39
Josh: say what you want, like they are in the lead.
23:41
Josh: It's maybe not the best piece of hardware in the world, but they have the best one available.
23:44
Josh: And we'll see what happens with the company. But that's kind of the update.
23:49
Josh: Meta is going open source again. The boys are back and they are spending a lot
23:53
Josh: of money and they are very clearly hellbent on making the best open source models
23:57
Josh: that they can and implementing them.
24:00
Josh: Across the entire product stack while continuing to iterate on their hardware
24:03
Josh: products and weird science projects like remember the the pupil dilation tracker
24:07
Josh: and like the this brainwave scanner that they're using to like understand how content affects you and
24:13
Josh: it's going to be a weird future with meta but they're working on it nonetheless
24:17
Josh: full speed ahead and um yeah we'll just continue to cover it as we go along this journey here
24:22
Ejaaz: Also i will just say this if you're listening to this episode and you are a
24:27
Ejaaz: builder, there has never been a better time to be alive.
24:31
Ejaaz: Like all these tools, all these models, all these products that are being released
24:35
Ejaaz: over the last couple of months that are being increased in frequency are all
24:39
Ejaaz: for you to trial test out.
24:41
Ejaaz: If it's too expensive for you to buy a Claude Mac subscription,
24:44
Ejaaz: play around with some of these open source models. If you don't think that the
24:47
Ejaaz: American frontier open source models are good enough, play around with some
24:49
Ejaaz: of these open weight or open source Chinese models.
24:52
Ejaaz: There is no excuse to not be using these things.
24:56
Ejaaz: If you have access to it, if you have a project or a hobby or an idea in your
25:00
Ejaaz: mind, go out there, try these things that we're talking about and honestly give
25:04
Ejaaz: us feedback because we'd love to hear from builders to see what you guys are
25:07
Ejaaz: actually doing with the thing.
25:08
Ejaaz: Maybe we're missing something and maybe you can kind of like shine a light on
25:11
Ejaaz: something that we're completely missing.
25:13
Ejaaz: But we will be covering anything and everything on Limitless going forwards.
25:17
Ejaaz: That is the end of this episode. You're all caught up on all of things meta
25:22
Ejaaz: related. They've done a pivot.
25:23
Ejaaz: They've turned it around and I'm excited to see where the company is in a couple months' time.
25:28
Ejaaz: That being said, wherever you're listening to us, if it's on YouTube and you
25:31
Ejaaz: aren't subscribed, if the notifications aren't turned on, please do so.
25:34
Ejaaz: It helps us out massively.
25:36
Ejaaz: If you haven't left us a comment on Spotify, Apple Music, or even YouTube as
25:39
Ejaaz: well, please do so. We have been loving the comments, by the way. I think we've
25:43
Ejaaz: gotten in the in the realm of like i don't know it's like a thousand two thousand
25:48
Ejaaz: plus comments over the last like five episodes keep them coming we love hearing
25:51
Ejaaz: from you guys we respond to anything and everything,
25:54
Ejaaz: as soon as we have like free time available um and then yeah we have a newsletter
25:57
Ejaaz: that goes out to over a hundred thousand of you we post twice a week i think
26:00
Ejaaz: this week we getting three times because josh and i can't keep our gobs shut
26:03
Ejaaz: but um josh anything anything well and most.
26:06
Josh: Most of all don't forget to gift this to a friend if there's someone else who
26:11
Josh: you know that would benefit from watching this every single day who wakes up
26:14
Josh: in the morning goes up with a little walk for their dog or like on their morning
26:17
Josh: commute that wants something to listen to 25 minutes long that will get sometimes
26:21
Josh: a little bit longer but that will get them up to speed on everything like
26:24
Josh: it's always much appreciated it's how we grow the best so as always thank you
26:27
Josh: so much for making it to the end of the episode real ones only here at the conclusion
26:31
Josh: thank you for watching and we will see you in the next episode