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

Auto-GPT: An Autonomous GPT-4 Experiment

github.com

Post reply on HN
Print view View original

141–150 of 178 posts

Re: Auto-GPT: An Autonomous GPT-4 Experiment

#141

Post by robwwilliams »

inciampati wrote: post #67
startupsfail wrote: post #55

Earlier quoted context omitted.

The problem is, human intelligence is likely also based on a similar advanced chat bot setup. While GPT-4 only performs as good as top-10th percentile of human students taking an exam (a professional in the field can do much more than this), it is notable that as a generalist GPT-4 would outperform such professionals. And GPT-4 is much faster than a human. And we have not yet evaluated GPT-4 working in its optimal se…

Alas, if it could only remember and precisely relate more than 4k or 8k or 32k or 64k words... And if only scaling that context length weren't quadratic... Indeed, we would really expect an AI to be able to achieve AGI. And it might decide to do all kinds of alien things. The sky would not be the limit! We have more than 100 trillion synapses in our brains. That's not our "parameter" count. It's the size of the thing…

> But these systems need to grow a ton of new helper functionality and subsystems to hope to achieve that kind of performance. And, I'll venture that kind of thing is a lower bound on the abilitites of any being who would be able to savage the world with it's intellect. It will have to be able to link up so, so many disparate threads to do it. It boggles our minds, which are only squaring a measly 100T dimension every tick.

Agreed: LLM are just one of many necessary modules. But amazing nonetheless. The quadratic scaling problem needs an attentional-conceptual extractor layer with working memory. Hofstadter points out that this needs to be structured as a recursive “strange loop” (p 709 of GEB). Thalamo-cortico-thalamic circuitry is a strange loop and attentional self-control may happens by phase- or time-shifting activity of different circuits to achieve flexible “binding” for attention and compute.

I’m actually optimistic that this is not a heavy computational lift but a clever deep extension of recursive self-modulating algorithms across modules. The recursion is key. And the embodiment is also probably crucial to bootstrap self-consciousness. Watching infants bootstrap is an inspiration.

Re: Auto-GPT: An Autonomous GPT-4 Experiment

#142

Post by tanseydavid »

ModernMech wrote: post #15
mrfinn wrote: post #13

Earlier quoted context omitted.

Wait for somebody mixing the concepts of "virus" and "AI". And we better begin to prepare some kind of AI sentinels because the question is not "if", but "when".

According to Agent Smith, virus + intelligence = humanity.

According to Bill Hicks, virus + shoes = humanity.

Re: Auto-GPT: An Autonomous GPT-4 Experiment

#143

Post by tanseydavid »

visarga wrote: post #28
jstanley wrote: post #11

Earlier quoted context omitted.

It doesn't. It makes the consequences more dramatic, if it (accidentally, even!) works out how to create its own successor, because at that point the genie is out of the bottle and you won't get it back in.

Needs a big honking datacenter or billions of compute credits and safety for 6-12 months.

Doesn't Alpaca seem to suggest that assumption is no longer true?

Re: Auto-GPT: An Autonomous GPT-4 Experiment

#144

Post by ALittleLight »

ModernMech wrote: post #139
ALittleLight wrote: post #137

Earlier quoted context omitted.

What do you call being able to play chess and play any other well known game and do well on a battery of standardized tests and write code in a variety of languages in a variety of problems and ask questions and write fiction prose or poetry and generally just take a shot at anything you happen to ask. I just can't take the idea that there is ambiguity as to whether these things have general problem solving skills se…

I would be able to make a long list of things while maintaining logical consistency with things earlier the list. For instance, I asked ChatGPT-4 to create a schedule for a class, and it started off okay, but by the time it got to the end of the schedule, it started listing topics already covered. Really shows how it's just going off of statistics.

This is an example of ChatGPT performing poorly, but not being unable to do the thing. Nobody would say would say ChatGPT has human level intelligence across all domains - but that it has general problem solving ability. In other words, I'm saying it has an IQ, not that it has the highest possible IQ.

And, of course, there are domains where ChatGPT will do better than you. Since I don't know your skill set I don't know what those domains are, but I assume you'd agree. Just like ChatGPT giving a bad schedule doesn't disprove it's intelligence, you not being able to come up with acrostics or pangrams easily (or whatever) doesn't disprove yours.

Re: Auto-GPT: An Autonomous GPT-4 Experiment

#145

Post by tanseydavid »

mjr00 wrote: post #56
ALittleLight wrote: post #45

Earlier quoted context omitted.

Language models obviously have some form of intelligence right now. You can have GPT-4 take SAT tests, play Chess, write poetry, predict what will happen in different social scenarios, answer theory of mind questions, ask questions, solve programming puzzles, etc. There are some measures that GPTs are clearly below human levels, some where they are far beyond, and some where they are within human range. The question…

> Language models obviously have some form of intelligence right now. This is not "obvious" in any sense of the word. At best, it's highly debatable.

>> This is not "obvious" in any sense of the word. At best, it's highly debatable.

Does a dog or cat have intelligence?

If you answered no, then I would ask if you don't you believe that by some measure a dog or cat has more intelligence than a rock?

And as a follow-on I would ask if you think GPT demonstrates more intelligence than a dog or a cat.

But perhaps you believe that in every one of these examples there is not a single case where it "obviously has some form of intelligence."

(I am really trying to highlight the semantic ambiguities)

Re: Auto-GPT: An Autonomous GPT-4 Experiment

#146

Post by tanseydavid »

roflyear wrote: post #133
ALittleLight wrote: post #45

Earlier quoted context omitted.

Language models obviously have some form of intelligence right now. You can have GPT-4 take SAT tests, play Chess, write poetry, predict what will happen in different social scenarios, answer theory of mind questions, ask questions, solve programming puzzles, etc. There are some measures that GPTs are clearly below human levels, some where they are far beyond, and some where they are within human range. The question…

Is stockfish intelligent?

>> Is stockfish intelligent?

It isn't general intelligence but I would argue that it is more intelligent than a new-born human being.

Re: Auto-GPT: An Autonomous GPT-4 Experiment

#147

Post by tanseydavid »

visarga wrote: post #30
rasengan wrote: post #8

That's pretty awesome. I wanted to try an experiment like that with ChatGPT moderating an IRC channel. I started here https://github.com/realrasengan/chatgpt-groupchat-test determining whether it was being spoken to and whether it should respond or not, and then providing that in json form so that it could be asked then to create the response in a sort of mini-chain of responses. It could also simply be asked, 'do yo…

Probably the only way to ensure civil forums in the future when nobody can be bothered to moderate by hand.

>> the only way to ensure civil forums

GPT will likely come to the conclusion that the only way to ensure civil forums is to keep the humans out entirely.

Re: Auto-GPT: An Autonomous GPT-4 Experiment

#148

Post by HyperSane »

krainboltgreene wrote: post #93
startupsfail wrote: post #55

Earlier quoted context omitted.

The problem is, human intelligence is likely also based on a similar advanced chat bot setup. While GPT-4 only performs as good as top-10th percentile of human students taking an exam (a professional in the field can do much more than this), it is notable that as a generalist GPT-4 would outperform such professionals. And GPT-4 is much faster than a human. And we have not yet evaluated GPT-4 working in its optimal se…

> The problem is, human intelligence is likely also based on a similar advanced chat bot setup. This is so wildly wrong and yet confidently said in every techbro post about LLMs. I beg of you to talk to an expert.

The fundamental truth is we really have no idea how the human brain works.

Re: Auto-GPT: An Autonomous GPT-4 Experiment

#149

Post by unsupp0rted »

boringuser2 wrote: post #60

I've been using GPT-4 extensively for programming and its consistent failures at novel tasks have kind of left me less excited.

It keeps getting stuff almost right and then when I make adjustments it fixes what I asked for but reverses previous things I asked for.

It’s like working on a project with an intermediate dev who keeps getting switched out for a brand new intermediate dev multiple times an hour.

Re: Auto-GPT: An Autonomous GPT-4 Experiment

#150

Post by chillfox »

inciampati wrote: post #67
startupsfail wrote: post #55

Earlier quoted context omitted.

The problem is, human intelligence is likely also based on a similar advanced chat bot setup. While GPT-4 only performs as good as top-10th percentile of human students taking an exam (a professional in the field can do much more than this), it is notable that as a generalist GPT-4 would outperform such professionals. And GPT-4 is much faster than a human. And we have not yet evaluated GPT-4 working in its optimal se…

Alas, if it could only remember and precisely relate more than 4k or 8k or 32k or 64k words... And if only scaling that context length weren't quadratic... Indeed, we would really expect an AI to be able to achieve AGI. And it might decide to do all kinds of alien things. The sky would not be the limit! We have more than 100 trillion synapses in our brains. That's not our "parameter" count. It's the size of the thing…

So how many of those 100 trillion synapses are actually in the part of the brain that does the thinking? Because the brain has different regions (subsystems) responsible for different things.
Post reply on HN