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

Voyager: An Open-Ended Embodied Agent with LLMs

voyager.minedojo.org

Post reply on HN
Print view View original

21–28 of 28 posts

Re: Voyager: An Open-Ended Embodied Agent with LLMs

#21

Post by usaar333 »

godelski wrote: post #19
usaar333 wrote: post #18

Earlier quoted context omitted.

> TLDR: see this prompt and ChatGPT's response And wow, that's GPT4. I've had similar thoughts as you. It feels like amazing intelligence one day, but the next seems like a extremely good, but naive pattern matcher. I've experienced similar GPT-4 disappoinments trying to teach it concepts not well in training data (it does badly) or making modifications to programs that go outside training data (e.g. make a tax calcu…

To be clear, that's ChatGPT, not GPT4. GPT4 should be better, but it is still limited beta and I haven't bothered joining. Note that 3.5-turbo (the API) is worse > They both weigh the same amount, which is 1 pound. It is clearly a strong example of Murry Gelman Amnesia when we can't trust it to tell us the difference between two simple things but we trust it to tell us complicated things. It is also a clear example o…

> GPT4 should be better, but it is still limited beta and I haven't bothered joining.

Ah my bad. Gpt4 via bing precise gets it correct:

> A kilogram of feathers weighs more than a pound of bricks. A kilogram is a metric unit of mass and is equivalent to 2.20462 pounds. So, a kilogram is heavier than a pound.

Re: Voyager: An Open-Ended Embodied Agent with LLMs

#22

Post by godelski »

YeGoblynQueenne wrote: post #20
godelski wrote: post #19

Earlier quoted context omitted.

To be clear, that's ChatGPT, not GPT4. GPT4 should be better, but it is still limited beta and I haven't bothered joining. Note that 3.5-turbo (the API) is worse > They both weigh the same amount, which is 1 pound. It is clearly a strong example of Murry Gelman Amnesia when we can't trust it to tell us the difference between two simple things but we trust it to tell us complicated things. It is also a clear example o…

>> To be clear, that's ChatGPT, not GPT4. GPT4 should be better, but it is still limited beta and I haven't bothered joining. Well, I can predict the next few token sequences you're about to get in response to your comment. "That's why you got that answer GPT4 is so much better" etc. Regarding your earlier comment about burnout, you're not alone. I stayed on HN because I could have the occasional good discussion abou…

Well it is good to know I'm not alone. These are strange times indeed. I often think one of the great filters of civilizations is overcoming a biological mechanism that designs brains to think simple (cheap compute/complexity is often unnecessary for survival objectives) and then advancing into a level of civilization where a significant amount of the problems the civilization require beyond first and second order approximations. (it happens when most challenges are solved to first and second order approximations) Unless one is able to rewire their consciousness I don't see how this wouldn't be a issue for any species but maybe I'm thinking too narrow or from too much of a bias.

Re: Voyager: An Open-Ended Embodied Agent with LLMs

#23

Post by osalberger »

Just checked by talking to the free version of ChatGPT, and yes, the MineFlayer api docs are indeed in its training set. It can give me detailed instructions on how to build a minecraft bot. And of course, it also knows the entire minecraft tech tree very well.

So this isn't really open ended work, its just making it do something it is already trained on, by connecting it to an API that it has learned the docs of.

Re: Voyager: An Open-Ended Embodied Agent with LLMs

#24

Post by osalberger »

osalberger wrote: post #23

Just checked by talking to the free version of ChatGPT, and yes, the MineFlayer api docs are indeed in its training set. It can give me detailed instructions on how to build a minecraft bot. And of course, it also knows the entire minecraft tech tree very well. So this isn't really open ended work, its just making it do something it is already trained on, by connecting it to an API that it has learned the docs of.

However, the skills library that it writes with live feedback from runtime errors and that it retrieves with a vector DB is really interesting. In that sense it looks like a very interesting code generation application.

Re: Voyager: An Open-Ended Embodied Agent with LLMs

#25

Post by FeepingCreature »

lsy wrote: post #12

I'm not sure how the authors arrive at the idea that this agent is embodied or open-ended. It is sending API calls to minecraft, there's no "body" involved except as a symbolic concept in a game engine, and the fact that minecraft is a video game with a limited variety of behaviors (and the authors give the GPT an "overarching goal" of novelty) precludes open-endedness. To me this feels like an example of the ludic f…

Though, I also don't individually track muscle fibers, and there's strong indications that a lot of my own behaviors are closer to API calls than direct control.

Re: Voyager: An Open-Ended Embodied Agent with LLMs

#27

Post by chinchilla2020 »

bartwr wrote: post #6

This is kind of amazing given that obviously, GPT-4 never contained such tasks and data. I think it puts an end to the claim that "language models are only stochastic parrots and cannot do any reasoning". No, this is 100% a form of reasoning and furthermore, learning that is more similar to how humans learn (gradient-less). I still don't understand it and it blows my mind - how such properties emerge just from compre…

I read through the code and tried it out for 15 mins.

It's a hard-coded program that can do a text search for it's own hard-coded, human-implemented functions. Apparently it can string those functions together, but doesn't do it correctly.

https://github.com/MineDojo/Voyager/tree/main/voyager/contro...

20 minutes of light reading through the repository pretty much dispels any notions that this is a self-learning system that can reason and think. It's the same minecraft automation we have been seeing for a decade now, with a chatbot text search builtin.

Re: Voyager: An Open-Ended Embodied Agent with LLMs

#28

Post by two_in_one »

godelski wrote: post #14
emptysongglass wrote: post #11

Earlier quoted context omitted.

You declare: > I think it puts an end to the claim that "language models are only stochastic parrots and cannot do any reasoning". But then two sentences later: > I still don't understand it and it blows my mind I've said this before to others and it bears repeating because your line of thinking is dangerous (not sudden AI cataclysm): to feel so totally qualified to make such a statement armed with ignorance, not kno…

> What is happening can be understood without resorting to the sort of magical thinking that ascribes agency to these models. This is what has (as an ML researcher) made me hate conversations around ML/AI recently. Honestly getting me burned out on an area of research I truly love and am passionate about. A lot of technical people openly and confidently are talking about magic. Talking as if the model didn't have acc…

> But it is still in line with the stochastic parrot paradigm.

What makes us different from 'stochastic parrots'? Or where creativity, which machines don't have by definition, begins and ends?

There is a bunch of philosophical questions, but LLMs are more than just parrots. They develop multi-level patterns recognition. And they can solve multi-step problems which they have never seen before. May be each individual step, but not the whole combination. Selecting the right combination out of zillons is not exactly 'parroting'. Doesn't matter how we call it, it has extremely high potential in real physical world. Looks like it's a near future.

We witness the emergency of 'Verbose AI'. IMHO. Which is more then just NLP

Post reply on HN