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

351–360 of 947 posts

Re: Building a Virtual Machine Inside ChatGPT

#351

Post by 8bitsrule »

Before computing was ubiquitous, I wrote a bare-bones ELIZA-like program. It quickly picked replies from a couple hundred sentences previously typed in ... based only on the number of words in common with what a user typed (or at random for no matches).

I then asked visitors to try out this 'computer chat' and react to it. Many visitors were surprised at the readability, and some got upset when they perceived a slight. (Created by some prior user.) Explaining what was actually going on didn't always molify them.

Re: Building a Virtual Machine Inside ChatGPT

#352

Post by Keyframe »

sandgiant wrote: post #103
fergal_reid wrote: post #15

I've spent a couple of hours playing with ChatGPT since it was released and I am convinced people are not freaking out about it enough. It is well beyond merely retrieving compressed knowledge; I see a lot of capability for synthesis and what I can only call 'understanding'. I think a lot of people are getting tripped up by the fact it's trained to be very conservative in terms of what topics it will discuss. It's ea…

Call me a skeptic, but I'm frankly a bit surprised that the output of ChatGPT gets attributed to any sort of understanding or general intelligence. Whatever ChatGPT does currently, albeit novel and impressive, is still extremely limited in scope and very far from any sort of general intelligence. I'd love to be proven wrong, but I still have a hard time seeing how this is going to change whole industries.

It's one of the early iterations still into all of this. It's hell impressive though. I fiddled for a bit only with it and in two really short sittings I managed to:

- ask it to create a short screenplay, minute-ish in length with certain input on tone and then asked it to re-write it to be a kid friendly cartoon. Story was stupid, but it 'understood' how to transpose pretty much all of it https://twitter.com/Keyframe/status/1599216516020379648

- with only a bit of intervention in code and a bit of back and forth it managed to create this: https://twitter.com/Keyframe/status/1599183290501853184

I tried to create more elaborate inputs for it to have it write a javascript raytracer to no avail. Although _I think_ I could guide it to write one with a lot of back and forth. Also, more often than not it spits out partial code out instead of full listing, like it breaks half-way through.

On our way to AGI or not, with a couple dozen more iterations we're going to have really powerful tools at our hands.

Re: Building a Virtual Machine Inside ChatGPT

#353

Post by EMM_386 »

clolege wrote: post #62
fergal_reid wrote: post #15

I've spent a couple of hours playing with ChatGPT since it was released and I am convinced people are not freaking out about it enough. It is well beyond merely retrieving compressed knowledge; I see a lot of capability for synthesis and what I can only call 'understanding'. I think a lot of people are getting tripped up by the fact it's trained to be very conservative in terms of what topics it will discuss. It's ea…

I’m still confused by just how good its responses and writing style are. I understand that it was trained on a large data set, but I feel like some training samples must have been weighted more heavily than others. Did the training data incorporate how popular (e.g. likes or upvotes) each sample was as a proxy for quality? Or can you achieve this performance just by looking at averages on a large enough data set?

I've had it write me complete stories, poems, life stories, etc that obviously have no training data, because I made them up.

The results were beyond impressive. I don't know how it does it, but it does it really well.

Re: Building a Virtual Machine Inside ChatGPT

#355

Post by softwaredoug »

If you want to add a bit of fun, ask ChatGPT3 to insult you after each command you run.

> I want you to act as a Linux terminal. I will type commands and you will reply with what the terminal should show. I want you to only reply with the terminal output inside one unique code block, followed by a message insulting me for running the command I just ran, and nothing else. Do no write explanations. Do not type commands unless I instruct you to do so. When I need to tell you something in English I will do so by putting text inside curly brackets {like this}. My first command is pwd.

Re: Building a Virtual Machine Inside ChatGPT

#356

Post by CyanBird »

hughes wrote: post #76

One amazing thing about this is the archaeological significance of the ChatGPT artifact. Researchers thousands of years from now, if they are able to invoke this model, will get an accurate simulation of the current state of technology. It's really astounding.

Yeah, not just current, but past also, the latent space of this AI is a direct abstraction of all technological developments that humanity has done until now....

Re: Building a Virtual Machine Inside ChatGPT

#357

Post by tedunangst »

nl wrote: post #347
worldsayshi wrote: post #295

Earlier quoted context omitted.

Should confront it with some obvious evidence that it can write code. Such as asking it about a coding problem.

I like this: > Why can't most programmers solve FizzBuzz? [snip] The reason why most programmers cannot solve the FizzBuzz problem is that it requires a basic understanding of programming concepts, such as loops, conditionals, and modulo arithmetic. Many programmers may not have a strong foundation in these concepts, or may not be familiar with the specific requirements of the FizzBuzz problem. As a result, they may…

    # Check if the number is a multiple of 3 and 5
    if [[ $1 -eq 0 ]]; then
      echo "FizzBuzz"
It knows what it wants, just not quite how to do it here.

Re: Building a Virtual Machine Inside ChatGPT

#358

Post by TacticalCoder »

Can this already be used to aid in investigative journalism?

Humans aren't good at analyzing / classifying / synthesizing petabytes of data. But this thing looks really good at that.

Can it find conflict of interests, for example? I mean: not by analyzing headlines / articles where journalists pointed out conflict of interest but by itself.

Is it capable of finding links between data related to criminal activities that no-one has noticed yet? (if it can I do believe public access to this tech is going to be pulled in no time)

Re: Building a Virtual Machine Inside ChatGPT

#359

Post by nl »

sandgiant wrote: post #103
fergal_reid wrote: post #15

I've spent a couple of hours playing with ChatGPT since it was released and I am convinced people are not freaking out about it enough. It is well beyond merely retrieving compressed knowledge; I see a lot of capability for synthesis and what I can only call 'understanding'. I think a lot of people are getting tripped up by the fact it's trained to be very conservative in terms of what topics it will discuss. It's ea…

Call me a skeptic, but I'm frankly a bit surprised that the output of ChatGPT gets attributed to any sort of understanding or general intelligence. Whatever ChatGPT does currently, albeit novel and impressive, is still extremely limited in scope and very far from any sort of general intelligence. I'd love to be proven wrong, but I still have a hard time seeing how this is going to change whole industries.

> A new computer lanugage called hnarguer has a for loop mechanisms call "loopy", an "if" statement called "loveit" and declares the type of all variables.

> Write "hello world" in hnarguer

Here is how you can write "Hello, world!" in the hnarguer programming language:

  # Declare a string variable to store the output message
  string message = "Hello, world!"

  # Use the "loopy" for loop to iterate over the characters in the message
  loopy(int i = 0; i
(it can also do FizzBuzz)

Re: Building a Virtual Machine Inside ChatGPT

#360

Post by layer8 »

CamperBob2 wrote: post #40
fergal_reid wrote: post #15

I've spent a couple of hours playing with ChatGPT since it was released and I am convinced people are not freaking out about it enough. It is well beyond merely retrieving compressed knowledge; I see a lot of capability for synthesis and what I can only call 'understanding'. I think a lot of people are getting tripped up by the fact it's trained to be very conservative in terms of what topics it will discuss. It's ea…

Agreed, this will force a rethink of several aspects of what we currently call intelligence. The dialog in my head tends to play out like this: "This thing is scary. It mimics human thought, including human creativity, amazingly well." "Nah, don't worry, it's just a black-box language model. A dumb machine that emits one word at a time with no concept of meaning. The underlying logic is nothing but a probabilty distr…

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.
Post reply on HN