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

291–300 of 622 posts

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

#291

Post by jedberg »

norir wrote: post #80
jedberg wrote: post #20

Earlier quoted context omitted.

Whether self awareness is a requirement for AGI definitely gets more into the Philosophy department than the Computer Science department. I'm not sure everyone even agrees on what AGI is, but a common test is "can it do what humans can". For example, in this article it says it can't do coding exercises outside the training set. That would definitely be on the "AGI checklist". Basically doing anything that is outside…

Here is an example of a task that I do not believe this generation of LLMs can ever do but that is possible for a human: design a Turing complete programming language that is both human and machine readable and implement a self hosted compiler in this language that self compiles on existing hardware faster than any known language implementation that also self compiles. Additionally, for any syntactically or semantica…

I will get excited when an LLM (or whatever technology is next) can solve tasks that 80%+ of adult humans can solve. Heck let's even say 80% of college graduates to make it harder.

Things like drive a car, fold laundry, run an errand, do some basic math.

You'll notice that two of those require some form of robot or mobility. I think that is key -- you can't have AGI without the ability to interact with the world in a way similar to most humans.

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

#292

Post by bob1029 »

norir wrote: post #80
jedberg wrote: post #20

Earlier quoted context omitted.

Whether self awareness is a requirement for AGI definitely gets more into the Philosophy department than the Computer Science department. I'm not sure everyone even agrees on what AGI is, but a common test is "can it do what humans can". For example, in this article it says it can't do coding exercises outside the training set. That would definitely be on the "AGI checklist". Basically doing anything that is outside…

Here is an example of a task that I do not believe this generation of LLMs can ever do but that is possible for a human: design a Turing complete programming language that is both human and machine readable and implement a self hosted compiler in this language that self compiles on existing hardware faster than any known language implementation that also self compiles. Additionally, for any syntactically or semantica…

This sounds like something more up the alley of linear genetic programming. There are some very interesting experiments out there that utilize UTMs (BrainFuck, Forth, et. al.) [0,1,2].

I've personally had some mild success getting these UTM variants to output their own children in a meta programming arrangement. The base program only has access to the valid instruction set of ~12 instructions per byte, while the task program has access to the full range of instructions and data per byte (256). By only training the base program, we reduce the search space by a very substantial factor. I think this would be similar to the idea of a self-hosted compiler, etc. I don't think there would be too much of a stretch to give it access to x86 instructions and a full VM once a certain amount of bootstrapping has been achieved.

[0]: https://arxiv.org/abs/2406.19108

[1]: https://github.com/kurtjd/brainfuck-evolved

[2]: https://news.ycombinator.com/item?id=36120286

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

#294

Post by olalonde »

rodgerd wrote: post #253
olalonde wrote: post #247

Earlier quoted context omitted.

I feel the test for AGI should be more like: "go find a job and earn money" or "start a profitable business" or "pick a bachelor degree and complete it", etc.

An LLM doing crypto spam/scamming has been making money by tricking Marc Andressen into boosting it. So to the degree that "scamming gullible billionaires and their fans" is a job, that's been done.

That story was a bit blown out of proportion. He gave a research grant to the bot's creator: https://x.com/pmarca/status/1846374466101944629

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

#295

Post by HarHarVeryFunny »

JumpCrisscross wrote: post #282
HarHarVeryFunny wrote: post #257

Earlier quoted context omitted.

It seems you are missing a lot of "ifs" in that hypothetical! Nobody knows how things like coding assistants or other AI applications will pan out. Maybe it'll be Oracle selling Meta-licenced solutions that gets the lion's share of the market. Maybe custom coding goes away for many business applications as off-the-shelf solutions get smarter. A future where all that AI (or some hypothetical AGI) changes is work being…

> you are missing a lot of "ifs" in that hypothetical The big one being I'm not assuming AGI. Low-level coding tasks, the kind frequently outsourced, are within the realm of being competitive with offshoring with known methods. My point is we don't need to assume AGI for these valuations to make sense.

Current AI coding assistants are best at writing functions or adding minor features to an existing code base. They are not agentic systems that can develop an entire solution from scratch given a specification, which in my experience is more typcical of the work that is being outsourced. AI is a tool, whose full-cycle productivity benefit seems questionable. It is not a replacement for a human.

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

#296

Post by darknoon »

devit wrote: post #286

It seems obvious to me that Common Crawl plus Github public repositories have more than an enough data to train an AI that is as good as any programmer (at tasks not requiring knowledge of non-public codebases or non-public domain knowledge). So the problem is more in the algorithm.

I think just reading the code wouldn't make you a good programmer, you'd need to "read" the anti-code, ie what doesn't work, by trial and error. Models overconfidence that their code will work often leads them to fail in practice.

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

#297

Post by eastbound »

ben_w wrote: post #208
whiplash451 wrote: post #197

Earlier quoted context omitted.

The main difference between GPT5 and a PhD-level new hire is that the new hire will autonomously go out, deliver and take on harder task with much fewer guidance than GPT5 will ever require. So much of human intelligence is about interacting with peers.

Human interaction with peers is also guidance. I don't know how many team meetings PhD students have, but I do know about software development jobs with 15 minute daily standups, and that length meeting at 120 words per minute for 5 days a week, 48 weeks per year of a 3 year PhD is 1.296.000 words.

I have 3 remote employees whose job is consistently as bad as LLM.

That means employees who use LLM are, on average, recognizably bad. Those who are good enough, are also good enough to write the code manually.

To the point I wonder whether this HN thread is generated by OpenAI, trying to create buzz around AI.

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

#298

Post by jedberg »

littlestymaar wrote: post #35
jedberg wrote: post #20

Earlier quoted context omitted.

Whether self awareness is a requirement for AGI definitely gets more into the Philosophy department than the Computer Science department. I'm not sure everyone even agrees on what AGI is, but a common test is "can it do what humans can". For example, in this article it says it can't do coding exercises outside the training set. That would definitely be on the "AGI checklist". Basically doing anything that is outside…

> Whether self awareness is a requirement for AGI definitely gets more into the Philosophy department than the Computer Science department. Depends on how you define “self awareness” but knowing that it doesn't know something instead of hallucinating a plausible-but-wrong is already self awareness of some kind. And it's both highly valuable and beyond current tech's capability.

When we test kids to see if they are gifted, one of the criteria is that they have the ability to say "I don't know".

That is definitely an ability that current LLMs lack.

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

#299

Post by kenjackson »

JohnMakin wrote: post #84
SpicyLemonZest wrote: post #48

Earlier quoted context omitted.

I don't understand why you'd be so dismissive about this. It's looking less likely that it'll end up happening, but is it any less believable than getting general intelligence by training a blob of meat?

> is it any less believable than getting general intelligence by training a blob of meat? Yes, because we understand the rough biological processes that cause this, and they are not remotely similar to this technology. We can also observe it. There is no evidence that current approaches can make LLM's achieve AGI, nor do we even know what processes would cause that.

> because we understand the rough biological processes that cause this

We don't have a rough understanding of the biological processes that cause this, unless you literally mean just the biological process and not how it actual impacts learning/intelligence.

There's no evidence that we (brains) have achieved AGI, unless you tautologically define AGI as our brains.

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

#300

Post by Timber-6539 »

Direct quote from the article: "The companies are facing several challenges. It’s become increasingly difficult to find new, untapped sources of high-quality, human-made training data that can be used to build more advanced AI systems."

The irony here is astounding.

Post reply on HN