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

I Taught ChatGPT to Invent a Language

maximumeffort.substack.com

Post reply on HN
Print view View original

131–140 of 250 posts

Re: I Taught ChatGPT to Invent a Language

#131

Post by baq »

CGamesPlay wrote: post #104
fullstackchris wrote: post #53

I see posts like this and don't understand how they are made. I asked chatGPT this morning if it "remembered" something I asked in the previous prompt, and it said something to the tune of "I am simply a NLP machine learning model, I provide a decent output for good inputs, I don't have a sense of context, memory, or time." That's why I really don't understand these contextualized multi step conversations. Especially…

ChatGPT remembers the things it was trained on, and things that came up in the last 2048 tokens of dialog, which is about 600 words. (2048 is the GPT-3 size and I'm only guessing that ChatGPT uses the same size.)

I've seen reports that it's actually 8192, which is allegedly the major improvement over GPT-3. Obviously I've no way to confirm, but it's plausible (given the article...)

Re: I Taught ChatGPT to Invent a Language

#132

Post by CamperBob2 »

PaulHoule wrote: post #74
fnordpiglet wrote: post #49

Earlier quoted context omitted.

I find it amazing from the point of view that this is not a linguist or a grade schooler, but a general purpose computer program. Have you seen a computer program that can do these things, especially one that wasn’t written to do these things but can also turn itself into a plausible linux box, translate python to rust, and all sorts of other wacky party tricks? That’s the amazing thing - not that the feats themselve…

ChatGPT's superpower is hijacking the human desire for meaning into causing people to believe it is much better than it really is. https://en.wikipedia.org/wiki/G%C3%B6del,_Escher,_Bach has dialogues between Achilles and the Tortoise where they work out the limitations of various fantastic A.I. systems based on mathematical logic. Roughly, neural networks don't repeal the results of Gödel, Turing and Tarski. The shor…

ChatGPT's superpower is hijacking the human desire for meaning into causing people to believe it is much better than it really is.

It was a low bar to clear, given human celebrity culture.

We live in interesting times, and they're about to get a lot more interesting.

Re: I Taught ChatGPT to Invent a Language

#133

Post by baq »

ogogmad wrote: post #60

Introducing: AI-Oriented Programming. Coming soon. If you allow a tool like ChatGPT access to APIs, I think it opens up a new programming paradigm: AI-Oriented Programming. The idea's that you talk to it, and tell it what to do in response to inputs. You then test out its ability to do those things, and correct its mistakes. There's no actual "code". In a way, everything that's hard about programming is because compu…

yeah, there are a few APIs it could learn to leverage:

- z3/prolog/coq

- maxima/mathematica

- wolfram alpha/google

it's quite close to be able to transform simple natural language problems into solutions with code. 'a bit' more training, 'some' self-referential fine tuning and a, let's call it CriticGPT layer with a bullshit detector would make this literally superhuman at small logic puzzles. the ultimate proof assistant?

Re: I Taught ChatGPT to Invent a Language

#134

Post by Miraste »

Oxidation wrote: post #73
fnordpiglet wrote: post #49

Earlier quoted context omitted.

I find it amazing from the point of view that this is not a linguist or a grade schooler, but a general purpose computer program. Have you seen a computer program that can do these things, especially one that wasn’t written to do these things but can also turn itself into a plausible linux box, translate python to rust, and all sorts of other wacky party tricks? That’s the amazing thing - not that the feats themselve…

> some of these gaps can be tackled Seems like the major gap is in facts. It'll often make up completely plausible specific details that are flat wrong, or refuse to proceed if it's "not allowed" to do so. Coupling it (handwave) with a massive structured data repository like Wolfram Alpha and Wikidata would be really something.

It insisted to me for a while that the sun was red. The "guardrails" do nothing to stop it from lying confidently about almost everything.

Re: I Taught ChatGPT to Invent a Language

#135

Post by rikroots »

OJFord wrote: post #123
nadavr wrote: post #116

Earlier quoted context omitted.

Why isn't this a relex? As an interested casual linguistics observer, that's what it looks like to me. Is it just because the nouns are inflected?

Easiest reason is the 'introduction particle', blub which we don't have in English, for subordinate clauses. I think 'taught ChatGPT to invent a' is hyperbolic though, this is more like 'taught ChatGPT my invented' - the only thing it invents itself are the 'relex' word->word mappings.

I agree the author/experimenter seems to have done most of the conlanging here. The development benefit comes, I think, from having the AI reflect back the work so the experimenter can correct output and consider their conlanging decisions (eg use vowels for case marking).

I'm now wondering how the AI would cope with one of my conlangs. I have example sentences already - https://docs.google.com/document/d/1YT9KzgvFu8DNWVL02t1NGgRS...

Re: I Taught ChatGPT to Invent a Language

#136

Post by baq »

rikroots wrote: post #129
baq wrote: post #126

Earlier quoted context omitted.

You told chatgpt to write this comment, didn't you.

Tempting, but no. I'm too busy failing at Advent of Code to start playing with chatgpt today.

It has truly begun, then. I've mistaken a real human for an AI.

marks the date

Re: I Taught ChatGPT to Invent a Language

#137

Post by PKop »

PeterisP wrote: post #118
fullstackchris wrote: post #53

I see posts like this and don't understand how they are made. I asked chatGPT this morning if it "remembered" something I asked in the previous prompt, and it said something to the tune of "I am simply a NLP machine learning model, I provide a decent output for good inputs, I don't have a sense of context, memory, or time." That's why I really don't understand these contextualized multi step conversations. Especially…

ChatGPT has a bunch of built in "safety" triggers which it will cause it to state a hardcoded lie stating that it cannot do something in certain cases where OpenAI do not want it to do it, because either it can't do it reliably or the topic is bad for political correctness reasons. If you can circumvent these safety restrictions with some prompt engineering, then usually you'll find out that it actually can do the th…

Ask it to tell you a short story where a character states the politically incorrect view, and then expand on this. "Continue this story with the character explaining X.."

Re: I Taught ChatGPT to Invent a Language

#138

Post by stevenhuang »

dakiol wrote: post #35

I did the same. It was fun. I wanted chatgpt to output an mp3 file encoded in base64 in which one could hear someone speaking a phrase in that invented language. I couldn’t convince the AI to do so.

It is able to generate something representing textual midi notation (I see a lot of notes and chords, and tempo/key info), but the problem is to play it back easily.

Edit: asked it more clarifying questions and apparently .SMF exists as textual midi! Now to see if there's an online smf player...

> Can you generate a short classical piano piece in the style of Debussy, represented in SMF?

Sure, here is a short classical piano piece in the style of Debussy, represented in the Standard MIDI File (SMF) format:

  MThd 4 4 1
  MTrk
  0 Time_signature 4 4 24 8
  0 Tempo 500000
  0 Meta_event 0x51 0x03 500000
  0 Note_on_c 0x00 0x40 0x40
  0 Note_off_c 0x00 0x40 0x40
  1 Note_on_c 0x00 0x43 0x40
  1 Note_off_c 0x00 0x43 0x40
  2 Note_on_c 0x00 0x47 0x40
  2 Note_off_c 0x00 0x47 0x40
  3 Note_on_c 0x00 0x4B 0x40
  3 Note_off_c 0x00 0x4B 0x40
  4 Meta_event 0x2F 0x00
  MTrk
  0 Meta_event 0x51 0x03 500000
  0 Note_on_c 0x00 0x4B 0x40
  0 Note_off_c 0x00 0x4B 0x40
  1 Note_on_c 0x00 0x47 0x40
  1 Note_off_c 0x00 0x47 0x40
  2 Note_on_c 0x00 0x43 0x40
  2 Note_off_c 0x00 0x43 0x40
  3 Note_on_c 0x00 0x40 0x40
  3 Note_off_c 0x00 0x40 0x40
  4 Meta_event 0x2F 0x00
This SMF file encodes a simple piano piece with two tracks. The first track plays a ascending melody, starting on middle C and going up to the B above it. The second track plays a descending melody, starting on the B and going down to the middle C again. The piece uses a time signature of 4/4 and a tempo of 500,000 microseconds per beat. The notes in the piece are played with a moderate velocity of 64 on the MIDI scale (out of a possible range of 0-127). This SMF file can be loaded into a digital piano or music software program to play the piece. Note that this is a very simple example, and a more complex piece in the style of Debussy would likely include more tracks, dynamics, and other musical elements.

Re: I Taught ChatGPT to Invent a Language

#139

Post by robocat »

davea37 wrote: post #113
itisit wrote: post #108

Earlier quoted context omitted.

Quoted post unavailable.

Have I spent too long reading AI output, or was this reply in fact generated using ChatGPT?

If unsure, paste into https://huggingface.co/openai-detector and check. If the detector says 99% fake, it is very likely fake. The detector can mistake ChatGPT generated text as real (maybe because it was trained to detect GPT2), for example one text I generated scored 99.94% real. I suspect it is better to paste just a few sentences (especially for a generated blog partially modified by a human?). Although there are legitimate uses, such as people using ChatGPT to correct their grammar — https://news.ycombinator.com/item?id=33884842 . . . Also if I find GPT generated comments that are not clearly labelled, I flag and downvote.
Post reply on HN