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

OpenAI, Google and Anthropic are struggling to build more advanced AI

bloomberg.com

Post reply on HN
Print view View original

601–610 of 622 posts

Re: OpenAI, Google and Anthropic are struggling to build more advanced AI

#601

Post by osigurdson »

jaculabilis wrote: post #580
visarga wrote: post #516

Earlier quoted context omitted.

You might personally have seen a deer just once, but human evolution, and animal evolution prior to that have practiced this skill a lot. AI doesn't have the advantage of evolutionary priors baked in, so it needs explicit walking through many combinations to infer its structure from data, and is remarkably efficient. GPT-4 'only' trained on the amount of language that 30,000 humans use in their lifetime. But we have…

> You might personally have seen a deer just once, but human evolution, and animal evolution prior to that have practiced this skill a lot. Which pre-human animals evolved instincts for swerving a car to avoid a deer?

I'm pretty sure that evolution would select out anything that could not generalize pretty quickly.

Re: OpenAI, Google and Anthropic are struggling to build more advanced AI

#602

Post by ppeetteerr »

jacobr1 wrote: post #576
netdevnet wrote: post #548

Earlier quoted context omitted.

Why don't you give actual concrete testable examples back with evidence where this is the case? Put your skin in the game.

A support ticket is a good middle ground. This is probably the area of most robust enterprise deployment. Synthesizing knowledge to produce a draft reply with some logic either to automatically send it or have human review. There are both shitty and ok systems that save real money with case deflection and even improved satisfaction rates. Partly this works because human responses can also suck, so you are raising a l…

Keyword is "draft". You still need a person to review the response with knowledge of the context of the issue. It's the same as my email example.

Re: OpenAI, Google and Anthropic are struggling to build more advanced AI

#603

Post by ppeetteerr »

jeswin wrote: post #482
ppeetteerr wrote: post #250

Earlier quoted context omitted.

The reason people are holding out is that the current generation of models are still pretty poor in many areas. You can have it craft an email, or to review your email, but I wouldn't trust an LLM with anything mission-critical. The accuracy of the generated output is too low be trusted in most practical applications.

Google (even now) wasn't absolutely accurate either. That didn't stop it from becoming many billions worth. > You can have it craft an email, or to review your email, but I wouldn't trust an LLM with anything mission-critical My point is that an entire world lies between these two extremes.

Google became a billion dollar company creating the best search and indexing service at the time and putting ads around the results (that and YouTube). The didn't own the answer of the question.

Re: OpenAI, Google and Anthropic are struggling to build more advanced AI

#604

Post by wslh »

tim333 wrote: post #593
wslh wrote: post #73

It sounds a bit sci-fi, but since these models are built on data generated by our civilization, I wonder if there's an epistemological bottleneck requiring smarter or more diverse individuals to produce richer data. This, in turn, could spark further breakthroughs in model development. Although these interactions with LLMs help address specific problems, truly complex issues remain beyond their current scope. With my…

AlphaGo which beat Lee Sedol was trained on human games. But then they produced AlphaZero which learned entirely from self play and got better than AlphaGo. So it goes.

That is just for chess which is not comparable to society/historical content, science, etc. Chess also have well defined rules.

Re: OpenAI, Google and Anthropic are struggling to build more advanced AI

#605

Post by salad-tycoon »

intended wrote: post #600
jacobr1 wrote: post #574

Earlier quoted context omitted.

We aren't talking about skilled knowledge work in Silicon Valley campuses. We are talking about work that might already have been outsourced so some cube-farm in the Philippines. Our routine office work that probably could already have been automated away by a line of business app in the 1980s, but is still done in some small office in Tulsa because it doesn't make sense to pay someone to write the code when 80% of t…

Ok. Well, I can see the direction you are going. I am unconvinced though - it hasn't thread the needle. Reason being 1) They are doing both in cube farms in the PHP, RTO + replacement by GenAI. 2) In high tech, they are also trying achieve these contradictory goals. RTO + Increased GenAI capability to reduce manpower needs. I can see a desire to reduce costs. I cant see how RTO to improve team work sits with using LL…

That’s a lot of weight on RTO and why it’s being implementing. A company is fully able to have you RTO, maybe even move, and fire you next day/month/year and desiring increased teamwork is not mutually exclusive of preparing for lay offs. Plus, I imagine at these companies there are multiple hands all doing things for their own purpose and metrics without knowing what the other hand is doing.Mid level Jan’s Christmas bonus depends on responding to exit interviews measurements showing workers leaving due to lack of teamwork, Bobs bonus depends on quickly implementing the code.

Re: OpenAI, Google and Anthropic are struggling to build more advanced AI

#606

Post by dmix »

ookdatnog wrote: post #533
dmix wrote: post #463

Earlier quoted context omitted.

Maybe in like 5yrs+. For now they will rake in billions just from API usage alone just with GPT4 and whatever 5 is. Amazon and Google didn't mess with their core business by competing with the players using it until they REALLY ran out of ways to make money.

OpenAI is losing far more billions than they are raking in. I don't think any generative AI company is even close to profitable at the moment. https://www.cnbc.com/2024/10/30/microsoft-cfo-says-openai-in...

When you make 4 billion in revenue you can generally figure out how to become profitable over time

High growth early days is a poor time to judge that

Re: OpenAI, Google and Anthropic are struggling to build more advanced AI

#607

Post by aniforprez »

danielbln wrote: post #583
namaria wrote: post #536

Earlier quoted context omitted.

It's not a gold standard. It just shows how difficult the problem really is. Flying machines rest on the excess power of internal combustion. They have nothing to do with bird evolution.

The fact that it has nothing to do with evolution is exactly my point. We built something that can fly but has nothing to do with how birds fly. So we might be able to build an AGI that isn't based on biological mechanism and/or evolutionary principles.

Planes don't fly radically differently than birds. Birds can flap their wings because they're light and small. Birds don't fly by flapping their wings, they flap their wings to fly. The flapping is to gain and maintain height but beyond that they use the same principle to stay afloat. Birds expend massive amounts of energy to flap too and eat a lot of food to compensate. Large predatory birds try their best to glide as much as possible as a consequence. To carry a human, you need a proportionally larger machine and the square-cubed law would stop us from being able to flap plane size wings. Aside from that, birds and planes fly on the same Bernoulli's Principle of fluid motion and to compensate for being unable to take off from rest with wings, we made engines that provide thrust.

If AGI doesn't take the form of human-ish intelligence, then we'd never know it was intelligence. This means that the target is always a "visible" human like intelligence and that was gained through evolution and millions of years of experimentation and records. It will most certainly not take that long for human-like intelligence to form given our current progress but we would not recognise anything else.

Re: OpenAI, Google and Anthropic are struggling to build more advanced AI

#608

Post by ben_w »

whiplash451 wrote: post #599
ben_w wrote: post #314

Earlier quoted context omitted.

1. The person I'm replying to is hypothesising about a future, not yet existent, version, GPT5. Current quality limits don't tell you jack about a hypothetical future, especially one that may not ever happen because money. 2. I'm not commenting on the quality, because they were writing about something that doesn't exist and therefore that's clearly just a given for the discussion. The only thing I was adding is that…

They definitely tell you jack. GPTs have reach their glass ceiling as they’ve sucked all available data and overfit to benchmarks. Their models have tons of use cases, but OpenAI and Anthropic are now in a product/commercial play.

That's one possibility.

Rumours have been in abundance since GPT-4 came out due to on the lack of clarity, but that lack of clarity seems to also exist within the companies themselves.

OpenAI and Anthropic certainly seem up be doing a lot of product stuff, but at the same time the only reason people have for saying OpenAI not making a profit is all the money they're also spending on training new models — I've yet to use o1, it's still in beta and is only 2 months old (how long was gmail in "beta", 5 years?)

I also don't know how much self-training they do, training on signals from the model's output and how users rate that output, only that (1) it's more then none, that (2) some models like Phi-3 use at least some synthetic data[0], and (3) that making a model to predict how users will rate the output was one of the previous big breakthroughs.

If they were to train on almost all their own output, and estimaing API costs as approximately actual costs, and given the claimed[1] public financial statements, that's in the order of a quadrillion (1e15) tokens, compared to the mere ~1e13 claimed for some of the larger models.

[0] https://arxiv.org/abs/2404.14219

[1] I've not found the official sources nor do I know where to look for them, all I see are news websites reporting on the numbers without giving citations I can chase up

Re: OpenAI, Google and Anthropic are struggling to build more advanced AI

#609

Post by BlueTemplar »

boredhedgehog wrote: post #537
sdesol wrote: post #248

Earlier quoted context omitted.

> "let's only allow the LLM to do things we know it is rock-solid at." Even this is insanely hard in my opinion. The one thing that you would assume LLM to excel at is spelling and grammar checking for the English language, but even the top model (GPT-4o) can be insanely stupid/unpredictable at times. Take the following example from my tool: https://app.gitsense.com/?doc=6c9bada92&model=GPT-4o&samples... 5 models are…

> I do believe LLM is a game changer, but I'm not convinced it is designed to be public-facing. I think that, too, is a UX problem. If you present the output as you do, as simple text on a screen, the average user will read it with the voice of an infallible Star Trek computer and be irritated by every mistake. But if you present the same thing as a bunch of cartoon characters talking to each other, users might not o…

> It looks like you're writing unsubstantiated nonsense. Would you like to turn it all caps ?

clippy.gif

Re: OpenAI, Google and Anthropic are struggling to build more advanced AI

#610

Post by tim333 »

irrational wrote: post #13

> The AGI bubble is bursting a little bit I'm surprised that any of these companies consider what they are working on to be Artificial General Intelligences. I'm probably wrong, but my impression was AGI meant the AI is self aware like a human. An LLM hardly seems like something that will lead to self-awareness.

Working towards it more than on it.

People use the term in different ways. It generally implies being able to think like a human or better. OpenAI have always said they are working towards it, I think deepmind too. It'll probably take more than an LLM.

It's economically a big deal because if it can out think humans you can set it to develop the next improved model and basically make humans redundant.

Post reply on HN