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

GPT-4 Designed a Programming Language

lukebechtel.com

Post reply on HN
Print view View original

221–230 of 338 posts

Re: GPT-4 Designed a Programming Language

#221

Post by geraneum »

eitland wrote: post #198
geraneum wrote: post #185

Earlier quoted context omitted.

> designing a language (albeit a derivative and uninspired one) is a trivial task that any human could achieve. It is, isn’t it? You just need to be able to read and write words and symbols that reassemble a programming language coherently and follow the pattern. This is what humans do after all; remixing. Anyone can do this.

Sorry if I am missing something here, but I think we all know well functioning, productive humans, maybe even in our field, who wouldn't have been able to create this language or a similarly well thought out one. I know a competent man should be able to "change a diaper, plan an invasion, butcher a hog, conn a ship, design a building, write a sonnet, balance accounts, build a wall, set a bone, comfort the dying, take…

I think you are missing my sarcasm. I don’t think every human can design a well thought out language, let alone an LLM. I was trying to point out that we are more than this. More than what humans are being demoted to in some HN comments recently. My problem is with some characterizations of humans here.

edit: typo fix

Re: GPT-4 Designed a Programming Language

#222

Post by tasuki »

> That's kind of neat -- what other even more creative ideas do you have for deeply integrating powerful LLMs directly into the language itself?

> [GPT lists 7 ideas]

My idea would be not to automate testing, but rather the implementation: the human writes the tests and the LLM writes the code to make them pass.

Automating testing is exactly the wrong way around: we let the LLM choose its own end goals. We'll be better off if we choose the end goals and task the LLM with achieving them.

Re: GPT-4 Designed a Programming Language

#223

Post by dgb23 »

yowzadave wrote: post #141

Last year, when Blake Lemoine was fired by Google for claiming his AI was sentient, a lot of people on this website sensibly noted that the danger of these chatbots was not that they were sentient, but that average people could easily be tricked into thinking that they were. And now here we are: it seems we tech people are every bit as susceptible to this kind of thinking as the average joes we looked down on. Actual…

I fully agree with your sentiment.

However I want to add that it does the guesswork and the presentation surprisingly well to a degree that it’s effective in certain contexts.

The more people think of this as a tool that they can use and master, the more they will benefit from it.

Re: GPT-4 Designed a Programming Language

#224

Post by testrun »

marviel wrote: post #138
testrun wrote: post #79

Earlier quoted context omitted.

Nope, we parrot mostly, but not always. When we do not, we create something new, and then we started parroting again. Mostly. The question is, can ai do something new, understand context of the new thing, understand when and where it will be usable and when it make sense to apply it.

When we're creating something "new" -- are you positing that it isn't built on top of other ideas, being combined in new ways?

New can be 1. something without a predecessor, 2. synthesis of other ideas that creates totally new application or theory

Re: GPT-4 Designed a Programming Language

#225

Post by amoss »

yowzadave wrote: post #141

Last year, when Blake Lemoine was fired by Google for claiming his AI was sentient, a lot of people on this website sensibly noted that the danger of these chatbots was not that they were sentient, but that average people could easily be tricked into thinking that they were. And now here we are: it seems we tech people are every bit as susceptible to this kind of thinking as the average joes we looked down on. Actual…

The interesting part of the exchange is the contrast between the surface level syntax of the exchange and the semantics below. It reminds me of teaching in a lot of ways. The level of polish on the immediately apparent surface level of the conversation is very high: reasonable tenents, good justification and ability to explain them.

But then applying them directly into an example falls down a bit. The language is a bit of a mismash and parts of it might not work semantically. The use of actors looks a bit out of place, and raises a lot of questions about how the memory management and scheduling is working under the hood. Despite the bold aims in the tenents thre is nothing in the language exposed so far that addresses them: why would the language produce more modular code in the large? what features of the syntax would increase the level of abstraction in code written? etc etc

Overall this would warrant a (barely) passing grade by modern standards, i.e. it can regurgitate the material succesfully but without demonstrating any deep understanding. I find the result quite interesting as it demonstrates how much of what we judge to be quality is actually fitting a particular pattern of response (w.r.t a specific field of study) rather than exploring the material more deeply.

I like the use of the arrows in the function definitions though :) I've seen that somewhere before and will be "appropriating" it for the project that I'm currently working on.

Re: GPT-4 Designed a Programming Language

#226

Post by dgb23 »

sbergot wrote: post #181
cyclotron3k wrote: post #153

Earlier quoted context omitted.

It seems only yesterday that we were marvelling at the capabilities of ChatGPT - it's ability to parse the sum knowledge of humankind and synthesise not trivial responses to questions we posed. But look how far we've come since those heady and carefree days. We're already able to pour buckets of scorn on its capabilities, and brush them aside as if designing a language (albeit a derivative and uninspired one) is a tr…

ChatGPT is both impressive and scary. We know how things work: there is no way to slow down this train. However it is already having an impact on things like education. Teacher don't have good enough tools to detect cheaters, and students are going to suffer because of that. Honest ones will be wrongly accused. Cheaters will get good grades. The only solution I see is to ban any graded homework. What will be the othe…

Students have been “cheating” with homework forever. Many get help from their parents and some even have professional help. Many collaborate or copy.

This is just another tool in the student’s toolbox and should be treated as such.

Re: GPT-4 Designed a Programming Language

#227

Post by marviel »

testrun wrote: post #224
marviel wrote: post #138

Earlier quoted context omitted.

When we're creating something "new" -- are you positing that it isn't built on top of other ideas, being combined in new ways?

New can be 1. something without a predecessor, 2. synthesis of other ideas that creates totally new application or theory

Can you provide an example of 1?

I think there are definitely some out there that feel almost "singular" in the way you describe, but each time I try to find a non-predecessor item I end up feeling "nope there are still dependencies"

Re: GPT-4 Designed a Programming Language

#228

Post by anonzzzies »

xlii wrote: post #170
yowzadave wrote: post #141

Last year, when Blake Lemoine was fired by Google for claiming his AI was sentient, a lot of people on this website sensibly noted that the danger of these chatbots was not that they were sentient, but that average people could easily be tricked into thinking that they were. And now here we are: it seems we tech people are every bit as susceptible to this kind of thinking as the average joes we looked down on. Actual…

Well put. I share the sentiment. I used GPT-3/3.5/4 for multiple tasks and I never found it useful. It just feels… average. It can “answer” many questions correctly but that’s only on the data set it’s been training on. I asked it to solve Elixir issues, write Elisp snippets, figure out why my Rust code was complaining about lifetimes and even went for easy mode and asked for JavaScript code. Only JavaScript (very po…

You probably never had to work with one of the 99% of programmers in the world that are not of your level (I am assuming it is high, otherwise you might just be very scared saying all this, and rightfully so), but me, working in the fortune 1000 company space, worked with many. Most of my colleagues can readily be replaced by gpt-4 today as far as coding and documentation goes. They make the same errors as you noticed gpt make (which I have to fix usually), only gpt does it in seconds vs hours and practically for free. By replacing all these people, I don’t have to build in the overhead of communicating with them (I just copy the jira task in gpt with some context and a few seconds later I have something better than most my colleagues write in 2-3 hours, on a good day), and they are happier doing nothing anyway. The company won’t fire them probably as managers get paid more and promoted more ‘managing’ more people.

This is not an exception either; teams like mine exist in every large enterprise. I worked in several of them and this AI makes life so much better. Let’s not pretend it makes certain outsourcing countries not fully obsolete though, to the tune of 10s of millions+ programmers, data entry workers, text producers (bloggers for seo, docs no one reads etc) and more are mostly all already worse than gpt3.5 and 4. Wait for 6 or 7.

Re: GPT-4 Designed a Programming Language

#229

Post by testrun »

vidarh wrote: post #178
testrun wrote: post #79

Earlier quoted context omitted.

Nope, we parrot mostly, but not always. When we do not, we create something new, and then we started parroting again. Mostly. The question is, can ai do something new, understand context of the new thing, understand when and where it will be usable and when it make sense to apply it.

Is something new that isn't a synthesis of past things not just a stochastic process? Human inventiveness seems to heavily follow a pattern of relatively minor iterations of what came before, extraordinarily rarely something that even seems to defy the past and be truly different and independent.

Democracy, nuclear power, calculus, dna helix, x-ray, written word, relativity comes to mind for starters. There are millions others.

Re: GPT-4 Designed a Programming Language

#230

Post by testrun »

marviel wrote: post #227
testrun wrote: post #224

Earlier quoted context omitted.

New can be 1. something without a predecessor, 2. synthesis of other ideas that creates totally new application or theory

Can you provide an example of 1? I think there are definitely some out there that feel almost "singular" in the way you describe, but each time I try to find a non-predecessor item I end up feeling "nope there are still dependencies"

Relativity for instance. New does not exclude dependencies or partial predecessors. New is something that is totally different from before.
Post reply on HN