Live data from Hacker News

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

nytimes.com

471–480 of 507 posts

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

#471
post #338

Earlier quoted context omitted.

> What is the VR adding? Nothing but extra hoops to just through to get things done. It gives you maximum immersion into a digital world. Rather than view it through a rectangular 2D window, it can encompass 360 degrees of your vision in full 3D. If you don't see how this would be appealing for consuming content, work, entertainment, etc., then I can't convince you otherwise. VR adoption has always been held back by…

> If you don't see how this would be appealing for consuming content, work, entertainment, etc., then I can't convince you otherwise. I don't, legitimately I don't. Okay, maximum immersion. And how does that help? Like even just on the surface having a 360 degree view doesn't do anything. Because my eyes are on the front of my head, so I'm going to be looking forward. Stuff behind me doesn't matter much. Same thing w…

What you're saying makes sense, and I do agree with some of it. But you're not seeing the experiences that a fully immersive digital world can deliver once the technology and UX improves.

Yes, we read and write on 2D surfaces, but a 3D environment allows you to have an infinite number of them. You could be writing on one, with a video playing in the background, while keeping an eye on a stock ticker, all inside a virtual beach, outer space, or whatever you want. You could be having a conversation with someone who looks and feels like they're physically in front of you. Have you seen the new visionOS avatars? They're incredibly realistic. Talking to a 2D video from a crappy webcam feels awfully primitive in comparison.

There are an infinite number of these experiences that we can't imagine yet, which a 2D display simply can't deliver.

As for HIDs, we'll figure it out. Voice and touch will become far more useful and user-friendly. Remember how touch keyboards on smartphones were unbearable to use initially, and tech geeks like myself strongly preferred a device with a physical keyboard? Well, text prediction and haptics got a lot better, and we invented swipe typing, so the experience improved considerably. I'm typing this on a swipe keyboard from my phone, and while I would prefer to use a keyboard on a real computer, I also like being able to type with one hand from bed. :)

So I'm sure we'll invent input mechanisms that feel natural and friendly to use in VR as well. In the meantime you can also use a physical keyboard. There are many software developers who have adopted AR glasses, with a phone and keyboard as their mobile workstations. The tech is almost there.

If you think about it, a 2D board with keys you press to compose words is an archaic method of inputting data into a machine. It's a remnant from typewriters dating back to the 19th century. They're far from being an optimal method for doing the type of things we use computers for today.

Ultimately, we can have different opinions on whether we want these experiences or not, but the reality is that this future is inevitable, for better or worse. As we move towards transhumanism, fully immersing our visual sense in a digital world is an obvious first step. The current iteration of devices that we're used to is merely ~50 years old, and ~20 for mobile devices. It's far too early in our technological evolution to consider these the best designs we can produce.

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

#472
post #436
post #411

Earlier quoted context omitted.

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?

i don't think anybody is having data labelers as employees.

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

#473
post #317

Earlier quoted context omitted.

If anything, "abstract links across domains" is the one area where even very low intelligence AI's will still have an edge, simply because any AI trained on general text has "learned" a whole lot of random knowledge about lots of different domains; more than any human could easily acquire. But again, this is true of AI's no matter how "smart" they are. Not related to any "super intelligence" specifically. Similarly,…

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 think they can make abstract links across domains.

Like the prompt "How can a simplicial complex be used in the creation of black metal guitar music?" https://chatgpt.com/share/684d52c0-bffc-8004-84ac-95d55f7bdc...

It is really more of a value judgement of the utility of the answer to a human.

Some kind of automated discovery across all domain pairs for something that a human finds utility in the answer seems almost like the definition of an intractable problem.

Superintelligence just seems like marketing to me in this context. As if AGI is so 2024.

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

#474
post #441
post #136

Earlier quoted context omitted.

Not a fan of Facebook or React (though React Native is IMHO the one eyed among the blind for cross platform mobile development), but I think that's a bit far fetched. I do think Facebook has (or had?) genuinely an engineering culture that wants to give something back.

So metaphorically speaking RN is an eye up over everything else?

From my personal experience, yes, but your mileage may vary. I tried to like Flutter multiple times, but can't. Capacitator etc feel fairly hacky. All the other options are comparatively obscure.

It helps that I like React alright. I don't think it's particularly well designed or enjoyable to use, and I hate debugging React code, but it's manageable with good development practices. Same goes for JS/TS. Not great, not terrible.

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

#475
post #361

Earlier quoted context omitted.

is he not? what about the whole metaverse thing and renaming the whole company to meta?

The metaverse will happen, IMO. The tech is just not there, yet. Even if it turns out to be wasted money, which I doubt, he's still sitting on almost 2 trillion. Not an L on my book.

Can you explain 2 trillion

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

#476
post #7

This is a weird deal? Meta buys a non-controlling stake and says no customers will be affected but the CEO and others are leaving Scale for Meta. Meta also says they won’t have access to competitor data but at 49% ownership they get major investor rights? Sounds like an acqui-kill to me?

Of all the tech companies, Meta is the most ruthless and shameless. You'd have to be a total fool to trust Zuck, especially Zuck who put billions into AR for not much return and now billions into AI to create lackluster lagging models.

[dead]

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

#477

Earlier quoted context omitted.

This is it. Maybe a few years ago at where I work, Oracle requires, as part of their licensing, the ability to scan every machine owned by the company to make sure there is no unlicensed use of any of their software. If any offending installations were found, they would charge the company the cost of the license for every machine. So, thousands of users times $thousands per license. Even if you had a license for a Ja…

OK, but that anecdote is orthogonal to your original claim. No mention of a lawsuit or actually having to pay extra fees. And "rogue installs" essentially means "using copyrighted software in a quantity that exceeds what we actually paid for", i.e. theft.

> No mention of ... actually having to pay extra fees.

> If any offending installations were found, they would charge the company the cost of the license for every machine.

I don’t think you read that correctly. To use an analogy, it’s similar to Costco requiring me to pay again for every item in my cart because I forgot to ring up a carton of eggs when I went through self-checkout. Maybe you don’t call that “extra” but I think most people would.

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

#478
post #85

The only way to understand this is by knowing: Meta already has two (!!) AI labs who are already at existential odds with one-another and both are in the process of failing spectacularly. One (FAIR) is lead by Rob Fergus (who? exactly!) because the previous lead quit. Relatively little gossip on that one other than top AI labs have their pick of outgoing talent. The other (GenAI) is lead by Ahmad Al-Dahle (who? exact…

I wouldn't categorize FAIR as failing. Their job is indeed fundamental research and are still a leading research lab, especially in perception and vision. See SAM2, DINOv2, V-JEPA-2, etc. The "fair" (hah) comparisons of FAIR are not to DeepMind/OAI/Anthropic, but to other publishing research labs like Google Research, NVIDIA Research, and they are doing great by that metric. It does seem that for whatever reason that…

FAIR is failing. Dino and JEPA at least are irrelevant in this age. This is why GenAI exists. GenAI took the good people, the money, the resources and the scope. Zuck tolerates entertains ideas until he doesn’t. It’s clear blue sky research is going to be pushed even further into the background. For perception reasons you can’t fire AI researchers or disband an ai research org but it’s clear which way this is headed.

As for your comparisons, well Google Research doesn’t exist anymore (to all intents and purposes) for similar reasons.

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

#479
post #56

Earlier quoted context omitted.

Of all the tech companies, Meta is the most ruthless and shameless. You'd have to be a total fool to trust Zuck, especially Zuck who put billions into AR for not much return and now billions into AI to create lackluster lagging models.

Off base when considering the likes of Palantir and many others. Not a fan of the person or many of Meta's business practices. But Meta has given a lot back with Llama and PyTorch, among many other open source contributions. Which others in the space are not doing.

I'm not sure their tech contributions in any way offset the various scandals they've had with data and like, actively targeting teens when it appears they're in a vulnerable state. Who cares about friggin 'open source' Llama when they've done evil stuff like that. Their 'social credit' is way deep in the red, esp with people who have no knowledge of those contributions.

They are incredibly ruthless and shameless in my opinion.

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

#480
post #95

Earlier quoted context omitted.

2 different topics, who you are (as a person, or company or other group of people) isn't some scale where you add + and - and judge final number. Bad/evil deeds define us 1000x more than positive ones and thats our lasting legacy and how we are/will be judged.

I agree with all of that. The point was to counter the statement that Meta is the most ruthless and shameless of all companies. The most ruthless and shameless company would not give back a lot of useful free open source software for hobbyists and companies to use.

Bread and circuses to keep the people happy while they do they shady stuff behind closed doors. This is millenia old. Can't say that as fact but it certainly looks that way when you examine their record.

Remember when they were into swaying elections, giving extra data to Cambridge analytica? Here's more

https://www.cnn.com/interactive/2018/12/business/facebooks-y...

Post reply on HN