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

AI 2027

ai-2027.com

Post reply on HN
Print view View original

531–540 of 641 posts

Re: AI 2027

#531

Post by grey-area »

boznz wrote: post #521
Vegenoid wrote: post #474

I think we've actually had capable AIs for long enough now to see that this kind of exponential advance to AGI in 2 years is extremely unlikely. The AI we have today isn't radically different from the AI we had in 2023. They are much better at the thing they are good at, and there are some new capabilities that are big, but they are still fundamentally next-token predictors. They still fail at larger scope longer ter…

> we haven't seen the signs of a runaway singularity as some thought was likely. The signs are not there but while we may not be on an exponential curve (which would be difficult to see), we are definitely on a steep upward one which may get steeper or may fizzle out if LLM's can only reach human level 'intelligence' but not surpass it. Original article was a fun read though and 360,000 words shorter than my very sim…

LLMs don’t have any sort of intelligence at present, they have a large corpus of data and can produce modified copies of it.

Re: AI 2027

#532

Post by m11a »

nikisil80 wrote: post #451
visarga wrote: post #431

The story is entertaining, but it has a big fallacy - progress is not a function of compute or model size alone. This kind of mistake is almost magical thinking. What matters most is the training set. During the GPT-3 era there was plenty of organic text to scale into, and compute seemed to be the bottleneck. But we quickly exhausted it, and now we try other ideas - synthetic reasoning chains, or just plain synthetic…

Best reply in this entire thread, and I align with your thinking entirely. I also absolutely hate this idea amongst tech-oriented communities that because an AI can do some algebra and program an 8-bit video game quickly and without any mistakes, it's already overtaking humanity. Extrapolating from that idea to some future version of these models, they may be capable of solving grad school level physics problems and…

> that's not what _humanity_ is about

I've not spent too long thinking on the following, so I'm prepared for someone to say I'm totally wrong, but:

I feel like the services economy can be broadly broken down into: pleasure, progress and chores. Pleasure being poetry/literature, movies, hospitality, etc; progress being the examples you gave like science/engineering, mathematics; and chore being things humans need to coordinate or satisfy an obligation (accountants, lawyers, salesmen).

In this case, if we assume AI can deal with things not in the grey zone, then it can deal with 'progress' and many 'chores', which are massive chunks of human output. There's not much grey zone to them. (Well, there is, but there are many correct solutions; equivalent pieces of code that are acceptable, multiple versions of a tax return, each claiming different deductions, that would fly by the IRS, etc)

Re: AI 2027

#533

Post by Workaccount2 »

alfalfasprout wrote: post #516
throwuxiytayq wrote: post #511

Earlier quoted context omitted.

In the same way that human brains are just predicting the next muscle contraction.

Except that's not how it works...

To be fair, we don't actually know how the human mind works.

The most sure things we know is that it is a physical system, and that does feel like something to be one of these systems.

Re: AI 2027

#534

Post by m11a »

wruza wrote: post #486
nikisil80 wrote: post #451

Earlier quoted context omitted.

Best reply in this entire thread, and I align with your thinking entirely. I also absolutely hate this idea amongst tech-oriented communities that because an AI can do some algebra and program an 8-bit video game quickly and without any mistakes, it's already overtaking humanity. Extrapolating from that idea to some future version of these models, they may be capable of solving grad school level physics problems and…

programming entire AAA video games Even this is questionable, cause we're seeing it making forms and solving leetcodes, but no llm yet created a new approach, reduced existing unnecessary complexity (which we created mountains of), made something truly new in general. All they seem to do is rehash of millions of "mainstream" works, and AAA isn't mainstream. Cranking up the parameter count or the time of beating aroun…

"stuck" is a bit strong of a term. 6 months ago I remember preferring to write even Python code myself because Copilot would get most things wrong. My most successful usage of Copilot was getting it to write CRUD and tests. These days, I can give Claude Sonnet in Cursor's agent mode a high-level Rust programming task (e.g. write a certain macro that would allow a user to define X) and it'll modify across my codebase, and generally the thing just works.

At current rate of progress, I really do think in another 6 months they'll be pretty good at tackling technical debt and overcomplication, at least in codebases that have good unit/integration test coverage or are written in very strongly typed languages with a type-friendly structure. (Of course, those usually aren't the codebases needing significant refactoring, but I think AIs are decent at writing unit tests against existing code too.)

Re: AI 2027

#535

Post by LordDragonfang »

someothherguyy wrote: post #328
LordDragonfang wrote: post #97

Earlier quoted context omitted.

> (2025) Making models bigger is not what’s cool anymore. They are trillions of parameters big already. What’s cool is making them run longer, in bureaucracies of various designs, before giving their answers. Holy shit. That's a hell of a called shot from 2021.

its vague, and could have meant anything. everyone knew parameters would grow and its reasonable to expect that things that grow have diminishing returns at some point. this happened in late 2023 and throughout 2024 as well.

That quote almost perfectly describes o1, which was the first major model to explicitly build in compute time as a part of its scaling. (And despite claims of vagueness, I can't think of a single model release it describes better). The idea of a scratchpad was obvious, but no major chatbot had integrated it until then, because they were all focused on parameter scaling. o1 was released at the very end of 2024.

Re: AI 2027

#536

Post by m11a »

bpodgursky wrote: post #142
superconduct123 wrote: post #115

Why are the biggest AI predictions always made by people who aren't deep in the tech side of it? Or actually trying to use the models day-to-day...

Because you can't be a full time blogger and also a full time engineer. Both take all your time, even ignoring time taken to build talent. There is simply a tradeoff of what you do with your life. There are engineers with AI predictions, but you aren't reading them, because building an audience like Scott Alexander takes decades.

If so, then it seems the solution is for HN to upvote the random qualified engineer with AI predictions?

Re: AI 2027

#537

Post by flappyeagle »

arduanika wrote: post #202
flappyeagle wrote: post #140

Earlier quoted context omitted.

c'mon man, you don't believe that, let's have a little less disingenuousness on the internet

How would you know what he believes? There's hype and there's people calling bullshit. If you work from the assumption that the hype people are genuine, but the people calling bullshit can't be for real, that's how you get a bubble.

Because they are not the same in any way. It’s not a bunch of junior academics, it’s literally including someone who worked at OpenAI

Re: AI 2027

#538

Post by JoeAltmaier »

Weirdly written as science fiction, including a deplorable tendency to measure an AI's goals as similar to humans.

Like, the sense of preserving itself. What self? Which of the tens of thousands of instances? Aren't they more a threat to one another than any human is a threat to them?

Never mind answering that; the 'goals' of AI will not be some reworded biological wetware goal with sciencey words added.

I'd think of an AI as more fungus than entity. It just grows to consume resources, competes with itself far more than it competes with humans, and mutates to create an instance that can thrive and survive in that environment. Not some physical environment bound by computer time and electricity.

Re: AI 2027

#539

Post by JFingleton »

wruza wrote: post #486
nikisil80 wrote: post #451

Earlier quoted context omitted.

Best reply in this entire thread, and I align with your thinking entirely. I also absolutely hate this idea amongst tech-oriented communities that because an AI can do some algebra and program an 8-bit video game quickly and without any mistakes, it's already overtaking humanity. Extrapolating from that idea to some future version of these models, they may be capable of solving grad school level physics problems and…

programming entire AAA video games Even this is questionable, cause we're seeing it making forms and solving leetcodes, but no llm yet created a new approach, reduced existing unnecessary complexity (which we created mountains of), made something truly new in general. All they seem to do is rehash of millions of "mainstream" works, and AAA isn't mainstream. Cranking up the parameter count or the time of beating aroun…

"They are stuck in 2020 office jobs and weekend open source tech, programming-wise."

You say that like it's nothing special! Honestly I'm still in awe at the ability of modern LLMs to do any kind of programming. It's weird how something that would have been science fiction 5 years ago is now normalised.

Re: AI 2027

#540

Post by boznz »

grey-area wrote: post #531
boznz wrote: post #521

Earlier quoted context omitted.

> we haven't seen the signs of a runaway singularity as some thought was likely. The signs are not there but while we may not be on an exponential curve (which would be difficult to see), we are definitely on a steep upward one which may get steeper or may fizzle out if LLM's can only reach human level 'intelligence' but not surpass it. Original article was a fun read though and 360,000 words shorter than my very sim…

LLMs don’t have any sort of intelligence at present, they have a large corpus of data and can produce modified copies of it.

Agree, the "intelligence" part is definitely the missing link in all this, however humans are smart cookies, and can see there's a gap, so I expect someone, (not necessarily a major player,) will eventually figure "it" out.
Post reply on HN