00:00So let's start with the big red number, and that is Meta's terrible monthly performance, I believe the worst showing
00:07since about October 2022.
00:09What do you make of the different headwinds hitting Meta right now and that big fear around just how much
00:14they're spending on AI?
00:16Yeah, look, I think everyone has been questioning their CapEx spend.
00:21You know, this year they're going to spend $135 billion, and they don't really have a model to show for
00:28that in terms of where frontier models are, you know, Gemini, Anthropic, and OpenAI.
00:35So Meta is nowhere on the leaderboard.
00:37You can't really see their token consumption, which is a key metric.
00:40And on top of that, they lost these, you know, two lawsuits where their platform safety is being questioned.
00:47And I expect maybe there will be more lawsuits because, you know, this is something that you're seeing regulators take
00:55a closer look at.
00:56It has been banned in Australia for under-16 users.
00:59So clearly there are things that are going against them when it comes to platform safety, and Section 230 Shield
01:07isn't really helping them here.
01:09Well, let's meditate a little bit longer on the spending point because when it comes to Meta, in some ways
01:14I'm having deja vu.
01:15You remember, you know, how much money they burned on the metaverse that led to their stock being down 64
01:21% in 2022.
01:24And then they had their year of efficiency where they said, we're going to really trim back, guys.
01:28And the stock was rewarded in a big way.
01:30And now you fast forward a couple of years, AI is the hot thing, and they're just shoveling money at
01:35it.
01:35Well, I mean, the good thing for someone like Meta is the core business is like a 60% plus
01:42incremental margin business.
01:44So they turn out so much free cash flow that they should be investing.
01:48The problem they have had is they spend $90 billion plus on reality labs and metaverse, and they have nothing
01:54to show for it.
01:55But hopefully that's not the case with, you know, LLMs because clearly everyone of their peers are investing big in
02:03AI infrastructure.
02:04It's just the monetization and return on LLM spend is not showing up yet.
02:11And that's where they've shown they are nimble.
02:13They can course correct quickly.
02:14So, you know, as long as the core business is strong, which it is, I mean, the addiction lawsuits kind
02:22of create that concern that maybe they may have to do more on platform safety.
02:26But, again, there's nowhere else that people are going.
02:30They're still spending, you know, a large portion of their time on Instagram and Meta apps.
02:35At what point, though, would you want to see AI agents start to pay dividends?
02:38Is this something that the market's willing to give them a year, six months, three months?
02:43Yeah, I mean, look, $135 billion in CapEx.
02:46If they really want to raise the CapEx for 2027, you want to show something that's comparable to, you know,
02:53what Claude has done or what OpenAI is doing on the enterprise side.
02:58And that's the other challenge for Meta.
03:00Meta is more focused on the consumer side.
03:02We know Google obviously has done quite well on the consumer side.
03:06So where does Meta fit in?
03:09And when things like this happen in terms of the lawsuit that they lost, it's very hard to create an
03:15ecosystem.
03:16With LLMs, it is about, one, having the model intelligence comparable to other models, but also creating an ecosystem.
03:24And I think that's where Meta's challenges are quite pronounced.
03:28Like, they can't really create an ecosystem when everyone is concerned about, you know, the platform being safe enough.
03:35In about 15 seconds, very quick here, reports coming out that they're testing subscriptions on Instagram.
03:40Is that a lever that the company realistically can pull?
03:43Yeah, I mean, the agentic distribution will happen through Instagram and Meta's family of apps.
03:49And they could roll out a subscription.
03:51The problem is they didn't really have a good product yet to, you know, roll that out through that distribution.
03:57So that's what I'm looking out for.
Comments