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

71–80 of 178 posts

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

#71

Post by pixl97 »

jliptzin wrote: post #49
sorokod wrote: post #12

AI needn't even try to escape, the clueless humans will plug it into reallity by themselves. Given this totally expected attitude I hope that the base models will never be released to the general population.

Can someone explain to me what is so dangerous here? Do people actually think we’re headed towards James Cameron’s terminator and soon ChatGPT will become self aware and destroy us? I am far more afraid of tools to edit viruses becoming cheap and widely available, then one nutjob in his garage can engineer and release smallpox v2.0 and wipe out 90% of the worldwide population before we even know what hit us. Or you k…

An AGI/ASI would be one of those tools that would make virus editing that much easier. One nutjob with a 'research scientist in a box' makes the drudgery of gene editing that much easier.

Now, if we're dumb enough to give AGI self motivation and access to tooling you can get paperclip maximizers, the AI could be the nutjob that you mention.

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

#72

Post by pixl97 »

jliptzin wrote: post #65
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…

But where is the imminent danger? It is still limited in many ways. For example, it can be turned off or unplugged. Is it because CAPTCHAs won’t work anymore? That sounds like a problem for sites like Twitter that have bot problems. Is it because it may replace people’s jobs? That comes with every technological step forward and there’s always alarmist ludditism to accompany it. Is it because bad people will use it to…

You didn't understand who the actual luddiets were, but don't worry, I have a feeling we'll get our chance.

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

#73

Post by flyinglizard »

rapsacnz wrote: post #66
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.

I've found this too. It confidently generates something for you... which turns out to be no good if it's in any way different from standard stuff. And making that leap from just a mash up of copy-pasta code to actual understanding... is huge. It wouldn't be an incremental upgrade, but a fundamental change in approach.

My experience it will just hallucinate what I'm after if I dare venture off the standard route. It's a rival for search, somewhat less so for expert humans.

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

#74

Post by startupsfail »

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…

You can only hold around 7 to 10 numbers in your mind well, in your working memory. Let me give you a few: 6398 5385 3854 8577

You have 1 second, close your eyes and add them together. Write down the result.

I’m pretty sure that GPT-4 at its 4k setting would outperform you.

[The point being, we have not seen what even GPT-4 can do in its optimal environment. Humans use paper, computers, google, etc. to organize their thoughts and work efficiently. They don’t just sit in empty space and then put everything into the working memory and magically produce the results. So imagine now that you do have a similar level of tooling and sophistication around GPT-4, like there is present around humans. I’m considering that and it is difficult to extrapolate what even GPT-4 can do, in its optimal environment.]

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

#75

Post by pixl97 »

29athrowaway wrote: post #36
sorokod wrote: post #12

AI needn't even try to escape, the clueless humans will plug it into reallity by themselves. Given this totally expected attitude I hope that the base models will never be released to the general population.

It is being used to suggest code completions, so it could suggest code completion to someone that upon execution becomes wormable malware that infects everyone.

Most companies at least still have tools that perform checks for at least simple versions of workable exploits. About the best chance it would have is to write a complex library and get everyone to use that, were the library is exploitable in a complicated fashion.

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

#76

Post by ALittleLight »

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.

I take intelligence to be a general problem solving ability. I think that's close to what most people mean by the term. By that definition it's clear that LLMs do have some level of intelligence - in some dimensions greater, lesser, or within the range of human intelligence.

What definition do you have for intelligence and how do LLMs fail to meet it?

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

#77

Post by cleanchit »

startupsfail wrote: post #43
smodad wrote: post #37

Earlier quoted context omitted.

I have to disagree. Not releasing it to the public makes it more dangerous. One major downside of all development being done in private is that the AI can very easily be co-opted by our self-appointed betters and you end up with a different kind of dystopia where every utterance, thought and act is recorded and monitored by an AI to make sure no one "steps out of line." I think the solution is releasing it to the gen…

I agree. Overall the whole situation feels like we’ve just entered atomic age and are proliferating plutonium, while selling shiny radioactive toys [I’m actually pretty serious here, the effects of prolonged interactions with an AI haven’t been evaluated yet, technically there is even a possibility of overriding a weak personality]. But it still feels like it is much safer to let GPT-4 loose and assess the consequenc…

Alright who plugged gpt into his hn account?

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

#78

Post by inciampati »

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 is simply unable to do anything novel. I've had arguments with friends about this, specifically in reference to the paper "Sparks of Artificial General Intelligence: Early experiments with GPT-4", which is wonderful and presents some amazing capabilities, many of which I use constantly for work every single day. But, these capabilities seem to all be within the range of data that it's trained on. Or they can be seen as interpolations, which are as novel as the prompter can suggest, but which are clearly derivatives of modes in the data and not of deep understanding of abstract concepts.

It's amazing stuff. But it totally fails to take the prompter anywhere new without extensive support, and it is still at a very shallow level of understanding with complex topics that require precision. For instance, turning a mathematical description of a completely novel (or just rare or unusual) algorithm into code will almost never work, and is more likely to generate a mess that takes lots of effort to clean up. And it's also extremely hard to get the model to self reflect and stop when it doesn't understand something. It is at present almost incapable of saying "I don't have enough information or structure to do X".

If we are already as deep into a realm of diminishing marginal returns as the GPT-4 white paper suggests, we might indeed be approaching a limit for this specific approach. No wonder someone is trying to dig a regulatory moat as fast as they can!

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

#79

Post by pixl97 »

mjr00 wrote: post #68
mythrwy wrote: post #62

Earlier quoted context omitted.

Would you settle for "behave exactly as if they had some form of intelligence"? Because from where I sit it's a distinction without a meaningful difference.

> Would you settle for "behave exactly as if they had some form of intelligence"? Sure, it behaves as if it has some form of intelligence in the sense that it can take external input, perform actions in reaction to this input, and produce outputs dependent on the input. This historically has been known as a computer program.

You're a computer program?

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

#80

Post by xg15 »

smodad wrote: post #37
sorokod wrote: post #12

AI needn't even try to escape, the clueless humans will plug it into reallity by themselves. Given this totally expected attitude I hope that the base models will never be released to the general population.

I have to disagree. Not releasing it to the public makes it more dangerous. One major downside of all development being done in private is that the AI can very easily be co-opted by our self-appointed betters and you end up with a different kind of dystopia where every utterance, thought and act is recorded and monitored by an AI to make sure no one "steps out of line." I think the solution is releasing it to the gen…

I get your point, but just to put it into perspective, you could theoretically use the same logic with bioweapons:

"Keeping this genetically engineered killer virus restricted to high security labs actually makes it more dangerous - it needs to be released into the wild, so people's immune systems have a chance to interact with the pathogen and develop natural immunity!"

Covid gave a taste how that kind of attitude would work out in practice.

Post reply on HN