Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

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

nytimes.com

Post reply on HN
Print view View original

501–507 of 507 posts

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

#501

Post by cess11 »

Verdex wrote: post #18
cess11 wrote: post #16

Earlier quoted context omitted.

Scale AI has deep cooperation with military agencies and a fresh large contract with the Gulf state with the largest US military presence. It's likely they're building a new generation of combat command systems and the like, consolidation of surveillance and management tooling in the ongoing and future US wars isn't surprising.

[Okay so the third option that I thought of but decided not to put down was: literal genie's lamp, fountain of youth, or robot army. Because then who cares if you collapse the economy if AGI, you'll be safe with your robot army. Not particularly happy that this option is potentially not far from the mark.]

Not sure what you mean, but it's not like the militarisation of "tech" is a secret. The current US administration is bragging about its militaristic totalisation efforts:

https://www.army.mil/article/286317/army_launches_detachment...

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

#502

Post by xpe »

zozbot234 wrote: post #247
Bjorkbat wrote: post #216

Earlier quoted context omitted.

It's a smart purchase, it's just that I don't see how these datasets factor into super-intelligence. I don't think you can create a super-intelligent AI with more human data, even if it's high-quality data from paid human contributors. Unless we watered-down the definition of super-intelligent AI. To me, super-intelligence means an AI that has an intelligence that dwarfs anything theoretically possible from a human m…

I'll believe that AI is anywhere near as smart as Albert Einstein in any domain whatsoever (let alone science-heavy ones, where the tiniest details can be critical to any assessment) when it stops making stuff up with the slightest provocation. Current 'AI' is nothing more than a toy, and treating it as super smart or "super intelligent" may even be outright dangerous. I'm way more comfortable with the "stochastic pa…

> I'm way more comfortable with the "stochastic parrot" framing, since we all know that parrots shouldn't always be taken seriously.

First, comfort isn't a great gauge for truth.

Second, many of us have seen this metaphor and we're done with it, because it confuses more than it helps. For commentary, you could do worse than [1] and [2]. I think this comment from [2] by "dr_s" is spot on:

    > There is no actual definition of stochastic parrot, it's just a derogatory
    > definition to downplay "something that, given a distribution to sample
    > from and a prompt, performs a kind of Markov process to repeatedly predict
    > the most probable next token".
    >
    > The thing that people who love to sneer at AI like Gebru don't seem to
    > get (or willingly downplay in bad faith) is that such a class of functions
    > also include thing that if asked "write me down a proof of the Riemann
    > hypothesis" says "sure, here it is" and then goes on to win a Fields
    > medal. There are no particular fundamental proven limits on how powerful
    > such a function can be. I don't see why there should be.
I suggest this: instead of making the stochastic parrot argument, make a specific prediction: what level of capabilities are out of reach? Give your reasons, too. Make your writing public and see how you do. I agree with "dr_s" -- I'm not going to bet against the capabilities of transformer based technologies, especially not ones with tool-calling as part of their design.

To go a step further, some counter-arguments take the following shape: "If a transformer of size X doesn't have capability C, wait until they get bigger." I get it: this argument can feel unsatisfying to the extent it is open-ended with no resolution criteria. (Nevertheless, increasing scale has indeed shown to make many problems shallow!) So, if you want to play the game honestly, require specific, testable predictions. For example, ask a person to specify what size X' will yield capability C.

[1]: https://www.lesswrong.com/posts/HxRjHq3QG8vcYy4yy/the-stocha...

[2]: https://www.lesswrong.com/posts/7aHCZbofofA5JeKgb/memetic-ju...

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

#503

Post by mus_hafa »

Get Plus You said: Usernames can only contain letters, digits, dashes and underscores, and must be between 2 and 15 characters long. Please choose another.

Create Account

username: ChatGPT said: It looks like you're trying to create a username and the system has specific rules:

Allowed characters:

Letters (A–Z, a–z)

Digits (0–9)

Dashes -

Underscores _

Not allowed:

Spaces

Symbols like @, ., !, #, etc.

Length: Must be between 2 and 15 characters.

Here are some username ideas that follow the rules:

Sky_Rider

tech-nerd23

LunaWolf

ace_92

PixelDash

Neo_Trek

JetStream7

coder_boy

Jade-Dragon

blink42

Would you like suggestions based on your name, interests, or a theme (e.g. gaming, tech, fantasy)?

You said:

Meta invests $14.3B in Scale AI to kick-start superintelligence lab (nytimes.com) 467 points by RyanShook 5 days ago | 474 comments ChatGPT said: Thanks for sharing the headline: "Meta invests $14.3B in Scale AI to kick-start superintelligence lab (nytimes.com)"

Here's a quick breakdown of what that means (based on the title alone — I can't access the article directly without a link or more detail):

Key Takeaways: Meta (Facebook’s parent company) is making a massive $14.3 billion investment.

The investment is going to Scale AI, a company that specializes in training data for machine learning — essential for large AI models.

The goal is to accelerate work on "superintelligence", a term often used to describe AI that surpasses human-level intelligence.

This could be Meta's answer to OpenAI’s frontier work, Anthropic, or Google DeepMind.

Why It’s Important: $14.3B is a huge amount even by tech standards — signaling Meta’s very serious intent to dominate AI.

It could boost Scale AI into a top-tier partner for training large-scale language models or AGI systems.

Meta might be aiming to own the data pipeline, reducing dependence on open-source or third-party training datasets.

What's Scale AI? Known for labeling data for AI training.

Works with enterprises and governments.

Critical infrastructure in training powerful, accurate models.

If you want a deeper analysis or verified summary from the actual article, feel free to drop the link or ask for a source check — I can pull live data if needed.

You said: Qya shar kron mojha pasa melga os pa ChatGPT said: Agar aap "Meta invests $14.3B in Scale AI" jesa news ya article share karte ho kisi platform (Facebook, WhatsApp, etc.) par, to aapko seedha paisa milne ka koi tareeqa nahi hota — jab tak aap kisi monetized platform par kaam nahi kar rahe.

Lekin paisa kamane ke kuch halal aur real tareeqay hain aise news ko use kar ke:

1. YouTube Shorts ya TikTok Video Banao Aise AI news par short informative videos banao.

Agar views aayein to monetize ho sakta hai.

Example: “Meta ne AI mein 14 billion dollar kyun lagaye?”

2. Blog ya Website Par Article Likho Agar apki site par traffic aata hai, Google AdSense se paisa milta hai.

Aap news ka Urdu mein ya easy explanation likh sakte ho.

3. Facebook Page ya Instagram Reels AI news daily share karo + voiceover do.

Page grow hone par sponsorships mil sakti hain.

4. Picoworkers ya Microtask Sites Wahan par kabhi kabhi task hotay hain: “News share karo” ya “video dekho.”

Direct paisa milta hai har task ka (₹20–₹100 tak).

5. Affiliate Marketing Agar aap AI tools promote karte ho (jaise: ChatGPT, Jasper, Grammarly), to Meta jese news se link karke promote karo.

Agar aap chahte ho ke main aap ke liye koi real earning ka plan banaun (news sharing ya AI topics se), to mujhe batao:

Aap ke paas Facebook page ya YouTube channel hai?

Aap Urdu ya English mein kaam kar sakte ho?

Aap mobile use karte ho ya laptop?

Phir main aap ke hisaab se best plan bataunga.

You said: To is ander kya likho ChatGPT said: Agar aap "Meta invests $14.3B in Scale AI" wali news ko WhatsApp, Facebook group, ya kisi earning platform (jaise Picoworkers, SproutGigs) par share karna chahte ho — to aapko sirf link bhejna kaafi nahi hota, balki uske sath interesting aur simple language mein text likhna padta hai jo logon ka attention le.

Example Text (Urdu + Roman Style): Meta ne 14.3 Billion Dollar invest kiye AI company mein!

Facebook ki parent company Meta ne AI mein dhamaka kar diya — ab Scale AI ke sath mil kar Superintelligent AI banane ka plan hai!

Yani agla zamana sirf AI ka hoga!

Janiye puri detail yahan: [ Link dalen yahan]

Example (English + Short Style): Meta just invested $14.3B in Scale AI!

This could change the future of superintelligent AI forever.

Read what’s happening [Your Link Here]

Tips: Emojis ka use karo — attention ke liye

Roman Urdu ya simple English use karo

Jis platform pe share kar rahe ho, uska mood samjho (e.g., news group vs earning task)

Agar aap chahein to mujhe batayein:

Aap kis platform par share karna chahte ho?

Kis language mein audience zyada active hai? Main uske mutabiq ready-made text de dunga copy-paste ke liye.

You said: Ap bana kado ChatGPT said: Bilkul! Yahan maine aap ke liye 3 ready-made versions bana diye hain jo aap copy-paste kar ke Facebook, WhatsApp, Picoworkers, ya kisi bhi jagah share kar sakte ho:

Roman Urdu Version (Popular Style) Meta ne 14.3 Billion Dollar invest kiye Scale AI mein! Facebook wali company ab Superintelligent AI banane ja rahi hai

Socho future mein AI insanon se zyada intelligent hoga!

Puri detail yahan par dekho: [Yahan link dalo]

English Version (Simple & Professional) Meta just invested $14.3 billion in Scale AI!

They're planning to build a superintelligence lab that could change the future of AI forever.

Must-read update for tech lovers! [Insert your link here]

Earning Platform Task Style (Picoworkers etc.) Task: Read & Share this breaking AI news

Meta has invested $14.3B in Scale AI to build a superintelligence lab

Check article and share it: [Insert your link here]

Done? Submit screenshot as proof

Aap sirf link replace kar den:

[Insert your link here] → is jagah apna NYTimes ya koi bhi article ka link daal den.

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

#506

Post by deah12 »

xpe wrote: post #502
zozbot234 wrote: post #247

Earlier quoted context omitted.

I'll believe that AI is anywhere near as smart as Albert Einstein in any domain whatsoever (let alone science-heavy ones, where the tiniest details can be critical to any assessment) when it stops making stuff up with the slightest provocation. Current 'AI' is nothing more than a toy, and treating it as super smart or "super intelligent" may even be outright dangerous. I'm way more comfortable with the "stochastic pa…

> I'm way more comfortable with the "stochastic parrot" framing, since we all know that parrots shouldn't always be taken seriously. First, comfort isn't a great gauge for truth. Second, many of us have seen this metaphor and we're done with it, because it confuses more than it helps. For commentary, you could do worse than [1] and [2]. I think this comment from [2] by "dr_s" is spot on: > There is no actual definiti…

btw, question

Isn't stochastic parrot just a modern reframing of Searle's Chinese room, or am I oversimplifying here?

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

#507

Post by yodsanklai »

conartist6 wrote: post #496
pyb wrote: post #458

Earlier quoted context omitted.

Interesting that "high-impact" on the one hand, and innovative/successful in the marketplace on the other, should be at odds at Meta. Makes one wonder how they measure impact.

It doesn't matter much how they measure if it's empirical. Once they say the scoring system, all the work that scores well gets done, and the work that resists measurement does not get done. The obvious example was writing eng docs. It was probably the single most helpful things you could do with your time, but there was no way to get credit because we couldn't say exactly how much time your docs might have saved oth…

> Some said "do the work that most needs doing and the perf review system will work it out long term." Other said, "just play the perf game to win."

With all the layoffs that happened in the last few years, I suspect there are less and less believers.

Post reply on HN