Meta invests $14.3B in Scale AI to kick-start superintelligence lab • Hacker News Board

Re: Meta invests $14.3B in Scale AI to kick-start superintelligence lab

#431

Post by psb217 » Sat, Jun 14, 2025, 1:20 AM UTC

CamperBob2 wrote: post #370

psb217 wrote: post #317

Earlier quoted context omitted.

I don't think current models are capable of making abstract links across domains. They can latch onto superficial similarities, but I have yet to see an instance of a model making an unexpected and useful analogy. It's a high bar, but I think that's fair for declaring superintelligence. In general, I agree that these models are in some sense extremely knowledgeable, which suggests they are ripe for producing producti…

I don't know about "useful" but this answer from o3-pro was nicely-inspired, I thought: https://chatgpt.com/share/684c805d-ef08-800b-b725-970561aaf5... I wonder if the comparison is actually original.

Comparing the process of research to tending a garden or raising children is fairly common. This is an iteration on that theme. One thing I find interesting about this analogy is that there's a strong sense of the model's autoregressiveness here in that the model commits early to the gardening analogy and then finds a way to make it work (more or less).

The sorts of useful analogies I was mostly talking about are those that appear in scientific research involving actionable technical details. Eg, diffusion models came about when folks with a background in statistical physics saw some connections between the math for variational autoencoders and the math for non-equilibrium thermodynamics. Guided by this connection, they decided to train models to generate data by learning to invert a diffusion process that gradually transforms complexly structured data into a much simpler distribution -- in this case, a basic multidimensional Gaussian.

I feel like these sorts of technical analogies are harder to stumble on than more common "linguistic" analogies. The latter can be useful tools for thinking, but tend to require some post-hoc interpretation and hand waving before they produce any actionable insight. The former are more direct bridges between domains that allow direct transfer of knowledge about one class of problems to another.

Re: Meta invests $14.3B in Scale AI to kick-start superintelligence lab

#432

Post by MichaelZuo » Sat, Jun 14, 2025, 1:33 AM UTC

jazzyjackson wrote: post #348

MichaelZuo wrote: post #234

Earlier quoted context omitted.

There are some stranger rumors too floating around.

This is America, if you have tea, spill it!

I’m nowhere near fully confident in these rumors… so there’s nothing to spill. I don’t post specific accusations without some completely reliable basis.

And I imagine that’s the norm in most places.

Re: Meta invests $14.3B in Scale AI to kick-start superintelligence lab

#433

Post by zozbot234 » Sat, Jun 14, 2025, 1:34 AM UTC

psb217 wrote: post #431

CamperBob2 wrote: post #370

Earlier quoted context omitted.

I don't know about "useful" but this answer from o3-pro was nicely-inspired, I thought: https://chatgpt.com/share/684c805d-ef08-800b-b725-970561aaf5... I wonder if the comparison is actually original.

Comparing the process of research to tending a garden or raising children is fairly common. This is an iteration on that theme. One thing I find interesting about this analogy is that there's a strong sense of the model's autoregressiveness here in that the model commits early to the gardening analogy and then finds a way to make it work (more or less). The sorts of useful analogies I was mostly talking about are tho…

> The sorts of useful analogies I was mostly talking about are those that appear in scientific research involving actionable technical details. Eg, diffusion models came about when folks with a background in statistical physics saw some connections between the math for variational autoencoders and the math for non-equilibrium thermodynamics.

These connections are all over the place but they tend to be obscured and disguised by gratuitous divergences in language and terminology across different communities. I think it remains to be seen if LLM's can be genuinely helpful here even though you are restricting to a rather narrow domain (math-heavy hard sciences) and one where human practitioners may well have the advantage. It's perhaps more likely that as formalization of math-heavy fields becomes more widespread, that these analogies will be routinely brought out as a matter of refactoring.

Re: Meta invests $14.3B in Scale AI to kick-start superintelligence lab

#434

Post by pests » Sat, Jun 14, 2025, 1:36 AM UTC

n2d4 wrote: post #304

gwbas1c wrote: post #273

Earlier quoted context omitted.

Because, in the 1990s, Pixar's IP was more popular than Disney's. The story (as I remember it from a Jobs biography) was that, in the Disney parks, there were longer lines for Pixar characters than Disney characters. Someone in leadership (don't remember the name) basically swallowed pride and bought Pixar from Jobs. It was considered a "reverse acquisition" because Jobs had so much stock he technically controlled Di…

This was certainly true for animated movies in the 2000s (where Pixar clearly dominated), although not the companies as a whole. Pixar shareholders (including Jobs) owned about 15% after the deal. This isn't a reverse acquisition, it's just a normal acquisition. Company A (Disney) has many things but is missing one thing (an animation team that doesn't suck), so they buy a company that does have that thing.

But what if company B ends up with majority ownership in company A in the end.

Re: Meta invests $14.3B in Scale AI to kick-start superintelligence lab

#435

Post by romanhn » Sat, Jun 14, 2025, 1:43 AM UTC

dongobread wrote: post #391

paxys wrote: post #146

Earlier quoted context omitted.

This is exactly why Zuck feels he needs a Sam Altman type in charge. They have the labs, the researchers, the GPUs, and unlimited cash to burn. Yet it takes more than all that to drive outcomes. Llama 4 is fine but still a distant 6th or 7th in the AI race. Everyone is too busy playing corporate politics. They need an outsider to come shake things up.

The corporate politics at Meta is the result of Zuck's own decisions. Even in big tech, Meta is (along with Amazon) rather famous for its highly political and backstabby culture. This is because these two companies have extremely performance-review oriented cultures where results need to be proven every quarter or you're grounds for laying off. Labs known for being innovative all share the same trait of allowing rese…

Can't upvote this enough. From what I saw at Meta, the idea of a high performance culture (which I generally don't have an issue with) found its ultimate form and became performance review culture. Almost every decision made filtered through "but how will this help me during the next review". If you ever wonder about some of the moves you see at Meta, perf review optimization was probably at the root of it.

Re: Meta invests $14.3B in Scale AI to kick-start superintelligence lab

#436

Post by dcsan » Sat, Jun 14, 2025, 1:45 AM UTC

trhway wrote: post #411

mattlondon wrote: post #200

Earlier quoted context omitted.

I wondered that. But then huge revenue streams for Scale basically disappear immediately. Is it worth Meta spending all that money just to stop competitors using Scale? There are competitors who I am sure would be very eager to get the money from Google, OpenAI, Anthropic etc that was previously going to Scale. So Meta spends all that money for basically nothing because the competitors will just fill the gap if Scale…

Meta buys 900 AI employees here at less than $20M/head. Pretty cheap these days. Any IP the company has is a bonus.

850 people doing data labelling?

Re: Meta invests $14.3B in Scale AI to kick-start superintelligence lab

#437

Post by lern_too_spel » Sat, Jun 14, 2025, 1:46 AM UTC

dongobread wrote: post #391

paxys wrote: post #146

Earlier quoted context omitted.

This is exactly why Zuck feels he needs a Sam Altman type in charge. They have the labs, the researchers, the GPUs, and unlimited cash to burn. Yet it takes more than all that to drive outcomes. Llama 4 is fine but still a distant 6th or 7th in the AI race. Everyone is too busy playing corporate politics. They need an outsider to come shake things up.

The corporate politics at Meta is the result of Zuck's own decisions. Even in big tech, Meta is (along with Amazon) rather famous for its highly political and backstabby culture. This is because these two companies have extremely performance-review oriented cultures where results need to be proven every quarter or you're grounds for laying off. Labs known for being innovative all share the same trait of allowing rese…

Beyond that, the leaders at Facebook are deeply unlikeable, well beyond the leaders at Google, which is not a low bar. I know more people who reflexively ignore Facebook recruiters than who ignore recruiters from any other company. With this announcement, they have found a way to make that problem even worse.

Re: Meta invests $14.3B in Scale AI to kick-start superintelligence lab

#438

Post by nrmitchi » Sat, Jun 14, 2025, 1:54 AM UTC

Wang has, seemingly, spent just as much time and energy over the last couple of years on PR stunts and publicity than on Scale itself. Between testifying to congress about how China is an AI risk (duh) and how AI is important (obviously), putting out press releases about joining boards of large orgs, and getting himself invited to trumps second inauguration. A high-billion-dollar story-headline framed as "Meta paying $14 BILLION for this one guy" is the same.

It very much seems it's been an investment in getting himself to be more of a "household name in AI". That is exactly what Meta needs (or at least thinks it needs) now.

I very much believe that there is very little moat in AI (currently, and in the forseeable future short some underlying hardware/etc breakthrough), and the success (from a consumer perspective) will come down to which of the big-cos (Facebook v Amazon v Google v OpenAI v Anthropic/Claude) consumers trust more. Zuck is, to put it mildly, *not* a trustworthy name for Meta to associate to leading the product that they want consumers to trust and depend on for their entire lives.

Whether or not Wang has any more qualifications than 1) is somewhat of a recognized AI name, and 2) is okay at speaking confidently on topics someone briefed him about, I don't think really plays much into this. If he needs help/assistance/etc with any of the meta scale/politics/management/etc, zuck can buy that for him.

What Zuck can't seem to buy (for himself) is some level of trust.

Re: Meta invests $14.3B in Scale AI to kick-start superintelligence lab

#439

Post by bigtones » Sat, Jun 14, 2025, 1:55 AM UTC

az226 wrote: post #398

nikcub wrote: post #383

Earlier quoted context omitted.

It's about 0.85% of Meta's market cap - less than the 1% they paid for (granted, all of) Instagram. They also paid about 1% of market cap for Oculus ($2b into a ~$220b market cap) Seems about par for Facebook when it comes to company-shifting acquisitions.

Little known secret is they paid $2.7B, not $2B. And Zuck and the FB head of M&A were talking shit about John Carmack’s crazy wife, who was doing his negotiation for him. On WhatsApp no less.

That sounds about right. Thanks for the insight.

Re: Meta invests $14.3B in Scale AI to kick-start superintelligence lab

#440

Post by nrmitchi » Sat, Jun 14, 2025, 2:00 AM UTC

tibbar wrote: post #380

molsongolden wrote: post #310

Earlier quoted context omitted.

Sounds like they're getting paid based on his note to employees: > "The proceeds from Meta's investment will be distributed to those of you who are shareholders and vested equity holders [...] The exceptional team here has been the key to our success, so I'm thrilled to be able to return the favor with this meaningful liquidity distribution." https://x.com/alexandr_wang/status/1933328165306577316

Yes, it is very good for employees and ex-employees.

Honestly if this acts as a liquidity event for a whole bunch of current employees, while at the same time giving off "Meta hand picked the CEO and whoever they felt were the best AI engineers and jumped ship" energy, I wouldn't be tooo surprised if current "scaliens" view this as the inflection point, and decide it's not worth staying for the other ~51% of their shares.