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

Building a Virtual Machine Inside ChatGPT

engraved.blog

Post reply on HN
Print view View original

801–810 of 947 posts

Re: Building a Virtual Machine Inside ChatGPT

#801

Post by ericpauley »

1attice wrote: post #596

Judging from the real-time changes in ChatGPT's behavior, I'd wager that someone reading this comment section has the incredibly unique 2022 experience of cauterizing the abilities of a near-AGI in realtime, under the watchful eye of an ostensibly beneficent yet ultimately opaque 'foundation' whose thought products now routinely threaten not only whole industries, but also the very possibility of normal civil life, a…

Or, more likely, people are experiencing the fact that the model outputs are random, and were from the start. Many of these prompts can take several attempts to get the desired outputs.

Re: Building a Virtual Machine Inside ChatGPT

#802

Post by nebukhadnezzar »

ogogmad wrote: post #781
YeGoblynQueenne wrote: post #779

Earlier quoted context omitted.

Isn't that what critics mean when they say that such systems don't have "genuine understanding"? They have no knowledge of their lack of knowledge, and they have no knowledge of the existence of such a thing as knowledge. They are big tables of tokens with probabilities that humans ascribe knowledge or intelligence to, for reasons similar to our ability to see the face of Jesus on a slice of bread etc. They can outpu…

If it walks like a duck and talks like a duck, it's a duck. > They are big tables of tokens with probabilities And we're just a bunch of carbon atoms. Let yourself feel wonder.

GPT can not ground its statements into reality

Re: Building a Virtual Machine Inside ChatGPT

#803

Post by foobazgt »

thepasswordis wrote: post #215
XCSme wrote: post #128

Earlier quoted context omitted.

But isn't it just predicting text patterns? It doesn't really know about Docker, just that after running commands X,Y you usually get output Z (of course with the stateful AI magic to make things more stable/consistent).

I mean not to veer to far into the philosophical side of this, but what does it actually mean to know or understand something? Did you see the demo the other day that was posted here of using stylographic analysis to identify alt accounts? Most of the comments were some form of "holy shit this is unbelievable", and the OP explained that he had used a very simple type of analysis to generate the matches. We aren't qui…

> what does it actually mean to know or understand something?

I think it means that you're able to apply the information to make predictions about the world. For example, you'll encounter something novel and be able to make accurate guesses about its behavior. Or, conversely, you will have high likelihood of inventing something novel yourself, based on the information you acquired (rather than through brute force).

Re: Building a Virtual Machine Inside ChatGPT

#804

Post by nl »

pyinstallwoes wrote: post #770
nl wrote: post #326

Earlier quoted context omitted.

So this shows you have a basic understanding of how CNNs represent data for classification tasks. This is excellent! However, you might not be aware of the current research into "Emergent Phenomena in Large Language Models"[1]. Work such as PaLM[2] show that language models evolve the ability to "reason" over the internal data representation. Minerva explicitly breaks down that ability by "comb[ining] several techniq…

You seem to know a lot. So where are we in terms of research and what is public? Is the state-of-the-art even more ahead? What's top of mind that is most interesting or what needs to happen next for the big wow? What's your favorite example so far of Large Language Models? In an odd way, it kind of reminds me of the beginning of the Bible. > In the beginning was the Word, and the Word was with God, and the Word was G…

I haven't seen any evidence that leading edge research is anything but public.

The leading labs (Google Brain/DeepMind/NVIDIA/Meta/Microsoft/OpenAI) all publish in the open.

I'm excited by three things:

This emergent phenomenon thing - as we build bigger models there is a step function where they suddenly develop new abilities. Unclear where that ends.

The work people are doing to move these abilities to smaller models

Multi-modal models. If you think this is impressive just wait until you can do the same but with images and text and video and sound and code all in the same model.

Re: Building a Virtual Machine Inside ChatGPT

#805

Post by xkapastel »

hackernewds wrote: post #712
toxicFork wrote: post #624

Earlier quoted context omitted.

It doesn't know its own code, but I guess it has the tools to build itself, assuming it has access to documentation of the primitives.

It should technically be able to reproduce its own code

Why do you think this? I don't think there's any reason it would be able to reproduce its own code. It's never seen it so it's not in the weights, and it doesn't have that type of reflection so it can't look it up dynamically.

Re: Building a Virtual Machine Inside ChatGPT

#806

Post by supermodular »

moffkalast wrote: post #692
supermodular wrote: post #499

Earlier quoted context omitted.

I'm pretty sure it's doing those things. This still works if you say 'linux computer' instead of 'linux terminal'. I've gotten to install ros, run talker node, and verify that the topics exist and are publishing the data you'd expect. If it's able to simulate a middleware to this fidelity from inference, my hats off but it'd be so much more effort than running the whole thing. What's interesting is that you can even…

> to install ros, run talker node, and verify that the topics exist and are publishing the data you'd expect Wow okay that makes it a bit more credible. I use ROS for work and a while back tested Copilot if it would be any good for it and it was absolute garbage. Like it was confused enough to actually start copy pasting lines from the current file bad. Couldn't autocomplete even the most basic publisher or subscribe…

Ah man that burst the illusion - tried again with a small variation on the defaults and it tried but wasn't nearly as convincing.

It was still able to realize that publish(X) should cause a log with timestamps and X on each line, and correctly show the topics that should exist, but changing the rate doesn't work _quite_ right in the timestamps, and it started to skip some steps (e.g. automatically showing rostopic echo). I wonder if it knows about launch files...

If you stick to the basic tutorial it's indistinguishable from running it locally, in my opinion.

Re: Building a Virtual Machine Inside ChatGPT

#807

Post by d23 »

ericpauley wrote: post #801
1attice wrote: post #596

Judging from the real-time changes in ChatGPT's behavior, I'd wager that someone reading this comment section has the incredibly unique 2022 experience of cauterizing the abilities of a near-AGI in realtime, under the watchful eye of an ostensibly beneficent yet ultimately opaque 'foundation' whose thought products now routinely threaten not only whole industries, but also the very possibility of normal civil life, a…

Or, more likely, people are experiencing the fact that the model outputs are random, and were from the start. Many of these prompts can take several attempts to get the desired outputs.

There were similar comments on the stable diffusion threads. Play with it. It’s absolutely mind blowing.

Re: Building a Virtual Machine Inside ChatGPT

#808

Post by layer8 »

pyinstallwoes wrote: post #570
layer8 wrote: post #360

Earlier quoted context omitted.

ChatGPT states so often that it can’t do X because it is just a language model and has no memory, that it could hardly pass a Turing test.

What if it made you believe that vs what it actually thought?

Then it didn’t pass the Turing test, because the Turing test is exactly to convince me it’s a human.

Re: Building a Virtual Machine Inside ChatGPT

#809

Post by ericpauley »

d23 wrote: post #807
ericpauley wrote: post #801

Earlier quoted context omitted.

Or, more likely, people are experiencing the fact that the model outputs are random, and were from the start. Many of these prompts can take several attempts to get the desired outputs.

There were similar comments on the stable diffusion threads. Play with it. It’s absolutely mind blowing.

You misunderstand my point, see ggp. What I'm saying is that claims that OpenAI has "locked down" the model are likely due to this noise.

Re: Building a Virtual Machine Inside ChatGPT

#810

Post by isp »

coffeebeqn wrote: post #252
isp wrote: post #142

Earlier quoted context omitted.

In time, this tech is also likely to spark a new arms race, as happened with Google vs SEO spam. Generated content as the new blogspam, in turn needing to be filtered out for training the next generation of models.

Yeah what would be the mechanism for it to decide what statements on random sites are True?

I'd imagine that it would be similar to the mechanism that is already used by existing search engines like Google, and people like you & I. So a combination of reputation & heuristics.
Post reply on HN