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

211–220 of 338 posts

Re: GPT-4 Designed a Programming Language

#211

Post by KyeRussell »

krupan wrote: post #163
marviel wrote: post #151

Earlier quoted context omitted.

I agree with your point about the dangers of chatbots being mistaken as sentient. And I also recognize the dangers of the self-improving systems like this in general -- that's why I'm not letting it run loose with a LangChainAI Agent script to self-execute a new language (though wouldn't that be fascinating?) While the language it generated needs refinement, the design process is iterative, and it takes time to devel…

Ok, but take a step back and really look at the whole situation here. So you've created a machine that solves problems in kind of the same way humans do. Yay? What overall problem is a machine like this solving? How is this better than just having a human do the job? It seems to have most of the same drawbacks as humans, and then some on top of that

What!? Look back at the last few hundred years, where the majority of great leaps have been the result of getting a machine to do something that humans did.

Re: GPT-4 Designed a Programming Language

#212

Post by sparkie »

Jensson wrote: post #208
sparkie wrote: post #202

Earlier quoted context omitted.

> It’s just a mishmash of features from the most commonly blogged-about existing programming languages. It has no conceptual insights, no original ideas, no taste that isn’t cribbed from one of the few languages it’s copying from. It will also learn all the bugs and reproduce them. Programmers who think ChatGPT is going to replace their job must have a very low opinion of their own skill set.

It will also combine correct things in incorrect ways, so even if everything it is trained on is correct it will still print incorrect things. We see it in this article, both the statement "Indentation-based scoping, similar to Python" and the code with bracers were probably correct in their original contexts, but combined they are no longer correct. LLMs will make such errors no matter what kind of data you train it…

Yes, and imagine if StackOverflow questions are part of the training material. The "AI" will be learning from unfinished/unworking code samples.

Re: GPT-4 Designed a Programming Language

#213

Post by avgcorrection »

zuminator wrote: post #190
avgcorrection wrote: post #164

Earlier quoted context omitted.

> 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. More. We’re more suspectible. We are the people who take the Turing Test as an objective test of GAI, after all, even though it only tests how we percieve supposed intelligence.

The Turing Test isn't meant to be an objective test of GAI. It's meant to replace a question that might be too vague or philosophically contentious to be answerable, "Can machines think?" with an answerable one, "Can a machine successfully fool a person into thinking it's a person?" So a robot bird that fools real birds arguably has passed the Avian Turing Test. (Also, arguably the Turing Test is by definition subjec…

> The Turing Test isn't meant to be an objective test of GAI. It's meant to replace a question that might be too vague or philosophically contentious [snip]

Correct in that that was what Turing intended. But the vulgar interpretation of it makes it out to be more than it really is.

For Pete’s sake: two years ago I even saw a psychologist who suggested that psychologists should take notes from how computer science ostensibly measures “intelligence” by the Turing Test, because it’s so “objective”. That was sad.

Re: GPT-4 Designed a Programming Language

#215

Post by GoblinSlayer »

raincole wrote: post #15
HervalFreire wrote: post #11

quotation from the link: "LLMs in the GPT family have read every programming language in the world, billions of times. It's been known for some time that these agents can write small snippets of code with some limited success. However, to my knowledge, none of these agents have ever been asked to make their own programming language. This type of task is a bit more complex, and requires a bit more creativity and fores…

I'm not so sure. The first sample is: > Indentation-based scoping, similar to Python. Then... > class Person { To me, it's very clear it doesn't really understand what "indentation-based scoping" means. For a human programmer, using both intentation-based and bracket-based scoping is a very unusual design choice, and is begging for further explanation. It talked pages of abstraction, modularity, etc, but not this ? I…

Not exactly this, but there's a human designed curly brace whitespace sensitive language. Whitespace sensitivity is a natural direction for syntax minimization.

Re: GPT-4 Designed a Programming Language

#216

Post by RHSman2 »

Jensson wrote: post #191
RHSman2 wrote: post #186

Earlier quoted context omitted.

The question is when does guessing become a moot point? AlphaGo guessed it’s way to being the best. Can something guess it’s way to evolving into a less guessy thing? I think so. Look at us? I am aware of how it works and used to be ‘it’s ok, it won’t happen’ but now I am actually pretty concerned. We think GAI as the thing but there maybe a reality that doesn’t require full GAI to completely unhinge us. I mean, what…

> AlphaGo guessed it’s way to being the best. No it didn't, AlphaGo was backed by an algorithm that perfectly understood go and used that to train itself to play go better. That is a completely different situation. If you tried to make a go engine that only tried to replicate old moves then you'd have a hard time making it perform valid moves. ChatGPT is like such an engine and we just made it perform valid moves mos…

You are too close.

AlphaGo was trained with tech. It beat humans.

ChatGPT et al are being trained.

Everyone is arguing the inner workings of what is in the box. It doesn’t matter.

The usage of the box(s) is what will be the big impact.

Previously those boxes couldn’t be relied upon so stayed down in the hierarchy of usefulness.

Now. They are rising.

Re: GPT-4 Designed a Programming Language

#217

Post by Jensson »

KyeRussell wrote: post #204
Jensson wrote: post #156

Earlier quoted context omitted.

> We're already able to pour buckets of scorn on its capabilities No we don't, we are just trying to put a damper on all the "the singularity is here!" style comments that seems to flood any discussion about GPT.

Oh please. This community is filled to the brim with people that are addicted to negativity and contrarianism. LLM criticisms very typically go far far beyond “this isn’t the singularity”, well toward “LLMs do nothing of value, snd certainly can’t I - the esteemed developer - with anything I find hard in my job!”

Not many say "LLMs do nothing of value". Many comments say LLMs aren't useful to them, but I haven't seen many say that LLMs aren't useful at all.

Re: GPT-4 Designed a Programming Language

#218

Post by hkt »

sparkie wrote: post #202
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…

> It’s just a mishmash of features from the most commonly blogged-about existing programming languages. It has no conceptual insights, no original ideas, no taste that isn’t cribbed from one of the few languages it’s copying from. It will also learn all the bugs and reproduce them. Programmers who think ChatGPT is going to replace their job must have a very low opinion of their own skill set.

> Programmers who think ChatGPT is going to replace their job must have a very low opinion of their own skill set.

As a devops guy I live in fear of anything that can copy and paste from stackoverflow faster than I can.

Re: GPT-4 Designed a Programming Language

#219

Post by seydor »

> Basically looks like Typescript and python had a baby.

And that's why it's useless. The model rehashed stackoverflow / hn /reddit from the past 10 years. There s no indication it's optimal, it's just (recently) popular

I don't know about you guys, but my excitement about GPT models is starting to deflate , i think they 've already beyond the peak of their abilities and more size is not enough for intelligence

Re: GPT-4 Designed a Programming Language

#220

Post by marviel »

ryanjshaw wrote: post #69
apatil wrote: post #38

It's looking increasingly possible that, at some point in the not-too-far future machines will be so good at creating software that humans won't be competitive in any way, and won't be in the loop at all. I happen to think that once machines reach this point humans won't be competitive in the labor market at all for long. It doesn't seem plausible that automatic driving would still be decades off, or that the trades…

This will almost certainly happen, and it will be a terrible mistake - at least in the story I'm working on :) My theory is that AI generated code will probably look and grow organically (the irony!). Humans will set out requirements, the AI will collate these into a series of tests, and it won't care how neat or understandable the code is, provided the tests pass. Basically an extremely diligent junior developer. Th…

I really hope we can figure it out. It's either we figure out alignment, we regulate ourselves VERY VERY CAREFULLY, or we perish -- I think?
Post reply on HN