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

40% of the code GitHub Copilot users check-in is AI generated and unmodified

microsoft.com

Post reply on HN
Print view View original

91–100 of 129 posts

Re: 40% of the code GitHub Copilot users check-in is AI generated and unmodified

#91

Post by tasuki »

worrycue wrote: post #79
tasuki wrote: post #77

Earlier quoted context omitted.

> thinking, designing and abstracting novel ideas into code is the hard part where AI would fall short and just waste my time How can you be sure of that? I know scientists who use ChatGPT to rewrite their text to be clearer. They've also used it to pose novel interesting questions about a situation. On my side, I've found ChatGPT very helpful to explore/compare which library to use to solve my problem and show examp…

Can you trust it though? It can’t really reason - it’s just very good at guessing the right answer if there are a lot of examples. It also lies somewhat frequently. If we have to double check its output, it kind of defeats the purpose - at least part of it - as the “cognitive burden” is still on us. P.S. Not sure what you mean by “lining up” Haskell types … but don’t LSPs already give us clear information about the i…

> Not sure what you mean by “lining up” Haskell types … but don’t LSPs already give us clear information about the inferred types and do so with nigh zero chance of errors.

I don't mean type inference but turning the present type a into the required type b. LSP don't do that.

Re: 40% of the code GitHub Copilot users check-in is AI generated and unmodified

#92

Post by TheUndead96 »

enlyth wrote: post #90
sph wrote: post #80

Earlier quoted context omitted.

AI is not as foolproof as many want us to believe. In my mind it's like giving a less experienced and error-prone engineer the code I'm working on to refactor. Perhaps it's because I've always been a lone wolf engineer, but I'd rather do it myself than giving quite an important task to someone less experienced than me, whether I'm writing something new (where I need to be at my best) or refactoring (where I still nee…

It's not foolproof but it does guess a lot of what you had exactly in your mind, just hadn't typed down yet. If I start "enum { North", I will gladly accept the ", East, South, West }" autocomplete from Copilot, and this kind of thing, in aggregate, saves hours of your time.

This is my feeling exactly, it is a semantic auto-complete. Should/can it write your core business logic? Probably not, but there is a good chance it can infer and suggest logical patterns in implementation.

Re: 40% of the code GitHub Copilot users check-in is AI generated and unmodified

#93

Post by tasuki »

js8 wrote: post #47
luxurytent wrote: post #5

I'm not surprised. A lot of the code I'll write in a day is trivial, calling helpers, writing test scaffolding, etc. I'm still building cool things the other 60% of time, but this job as we know it comes with a lot of repetition. (esp. if you write unit tests ;-))

The real question is, why we need to write "trivial" code? I think people have lost the plot there. We should be writing better compilers and libraries, instead of using AI that will write more code to deal with. The obsession with code generation instead of creating proper abstractions is a scourge of today's programming. It's a regression in software engineering.

I couldn't agree more with everything you wrote. The problem with repetition is not that it's hard to generate, the problem is it's hard to maintain!

I've used ChatGPT to find libraries and to help me line up types. It was very helpful.

Re: 40% of the code GitHub Copilot users check-in is AI generated and unmodified

#94

Post by famahar »

ok_dad wrote: post #67

So let’s have a test where two or more teams build the same software. One set has AI tools, the other doesn’t. The software definition is provided to both at the same time, and there is a set of tests to see which team produced the software the fastest and produced the highest quality software.

I'd watch that game show.

Re: 40% of the code GitHub Copilot users check-in is AI generated and unmodified

#95

Post by lionkor »

tasuki wrote: post #89
worrycue wrote: post #79

Earlier quoted context omitted.

Can you trust it though? It can’t really reason - it’s just very good at guessing the right answer if there are a lot of examples. It also lies somewhat frequently. If we have to double check its output, it kind of defeats the purpose - at least part of it - as the “cognitive burden” is still on us. P.S. Not sure what you mean by “lining up” Haskell types … but don’t LSPs already give us clear information about the i…

> Can you trust it though? Yes of course. Libraries: I look at the libraries it suggests, see if they have documentation or GitHub activity. Lining up types: say you have a `List (Maybe a)` and require a `Maybe (List a)`. This is a simple example, of course real life is much more complicated. ChatGPT tells you. Can you trust it? Well did it make the compile error go away? (Sure, I can think of three ways to turn a `L…

No, you can't trust it. It will certainly write bugs, and the code will have other issues. Usually too subtle for you to see at first glance, like off-by-one errors, or uncovered edge cases.

Re: 40% of the code GitHub Copilot users check-in is AI generated and unmodified

#96

Post by euroderf »

ok_dad wrote: post #67

So let’s have a test where two or more teams build the same software. One set has AI tools, the other doesn’t. The software definition is provided to both at the same time, and there is a set of tests to see which team produced the software the fastest and produced the highest quality software.

Call it The 80-20 Show. Compared to the not-AI'd team, the AI'd team gets 80% of the correctness in 20% of the time. Then it is revealed to be an application where errors are life-threatening. Thrills! Dread! FUN!

Re: 40% of the code GitHub Copilot users check-in is AI generated and unmodified

#97

Post by never_inline »

js8 wrote: post #86
never_inline wrote: post #62

Earlier quoted context omitted.

I have made the same point elsewhere to AI enthusiasts friends of mine. The fact that defining the same task in natural language to an AI is easier than writing it in formal language, means the formal language lacks expressiveness [0], vocabulary, or more likely, both. But if they are both at same level of expressiveness and vocabulary, formal language wins, because it can spot and prevent ambiguity. The question is…

> How many languages allow to straightforwardly express `clamp m between 1.0 and 2.0`? You can argue this is vocabulary, or this is expressiveness. And I think it would be a great place for AI in programming - automated refactoring. For example, it would see a code that clamps the value and could be replaced with a library function call clamp(m,1,2). (And it would also, presumably, made sure that the code is indeed t…

Good point. But I think current approach of large scale generative AI models is not sufficient for this.

Re: 40% of the code GitHub Copilot users check-in is AI generated and unmodified

#98

Post by KyeRussell »

sph wrote: post #69

Aside: am I the only one here that is still writing code by hand and not using AI to autogenerate it? Because it seems over the past couple years on HN everybody is making great use of Copilot and constantly talking about it, while honestly scaffolding boilerplate is the least of my problems, while thinking, designing and abstracting novel ideas into code is the hard part where AI would fall short and just waste my t…

Weird flex or whatever dude. But all I’m hearing is that your anxiety has led you to a) refuse to look into a new piece of technology, b) hold incorrect views on it (like that it’s only useful for boilerplate code), c) look down on those that ARE using it, and d) pose that your incredible experience is why this wouldn’t be useful to you.

If you can’t find your drive to look at new things enough to even see if they’re worth digging more into, which you seemingly haven’t, then I think this speaks poorly of you as a developer.

Re: 40% of the code GitHub Copilot users check-in is AI generated and unmodified

#99

Post by KyeRussell »

sph wrote: post #80
ramblerman wrote: post #78

Earlier quoted context omitted.

> while thinking, designing and abstracting novel ideas into code is the hard part where AI would fall short and just waste my time. So don't use it there :) But I'm curious what kind of coding you do that you never have to do any scaffolding, refactoring, or come across any element of repetitiveness. That is where it shines imo. I find during those boring tasks it quite often just gets my intent, and I can say "yes…

AI is not as foolproof as many want us to believe. In my mind it's like giving a less experienced and error-prone engineer the code I'm working on to refactor. Perhaps it's because I've always been a lone wolf engineer, but I'd rather do it myself than giving quite an important task to someone less experienced than me, whether I'm writing something new (where I need to be at my best) or refactoring (where I still nee…

Again, you’ve basically foregone empiricism and think that the Internet needs your uninformed reckonings about the shortcoming of a technology that you seemingly haven’t had experience with.

It sounds like you’ve been at this for a while, are set in your ways, and are trying to justify it. We all get set in our ways, but stop trying to pass this off as something it’s not.

I’d respect using Copilot and deciding that it didn’t provide value, but all this umming and ahhing is not that.

Re: 40% of the code GitHub Copilot users check-in is AI generated and unmodified

#100

Post by sph »

KyeRussell wrote: post #98
sph wrote: post #69

Aside: am I the only one here that is still writing code by hand and not using AI to autogenerate it? Because it seems over the past couple years on HN everybody is making great use of Copilot and constantly talking about it, while honestly scaffolding boilerplate is the least of my problems, while thinking, designing and abstracting novel ideas into code is the hard part where AI would fall short and just waste my t…

Weird flex or whatever dude. But all I’m hearing is that your anxiety has led you to a) refuse to look into a new piece of technology, b) hold incorrect views on it (like that it’s only useful for boilerplate code), c) look down on those that ARE using it, and d) pose that your incredible experience is why this wouldn’t be useful to you. If you can’t find your drive to look at new things enough to even see if they’re…

I'm sorry, but I do not get why you're making it personal and attacking my worth as a developer because I am doubtful about what Copilot can do for me.

It is possible to disagree without being rude and insulting.

Post reply on HN