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

Using LLMs at Oxide

rfd.shared.oxide.computer

Post reply on HN
Print view View original

151–160 of 284 posts

Re: Using LLMs at Oxide

#151

Post by bananapub »

philippta wrote: post #145

> LLM-generated code should not be reviewed by others if the responsible engineer has not themselves reviewed it. To extend that: If the LLM is the author and the responsible engineer is the genuine first reviewer, do you need a second engineer at all? Typically in my experience one review is enough.

yes, obviously?

anyone who is doing serious enough engineering that they have the rule of "one human writes, one human reviews" wants two humans to actually put careful thought in to a thing, and only one of them is deeply incentivised to just commit the code.

your suggestion means less review and worse incentives.

Re: Using LLMs at Oxide

#152

Post by Madmallard »

"LLMs can be quite effective writing code de novo."

Maybe for simple braindead tasks you can do yourself anyway.

Try doing it on something actually hard or complex and they get it wrong 100/100 if they don't have adequate training data, and 90/100 if they do.

Re: Using LLMs at Oxide

#153

Post by menaerus »

Barrin92 wrote: post #88
keyle wrote: post #44

Earlier quoted context omitted.

> That junior engineer possibly hasn't programmed without the tantalizing, even desperately tempting option to be assisted by an LLM. This gives me somewhat of a knee jerk reaction. When I started programming professionally in the 90s, the internet came of age and I remember being told "in my days, we had books and we remembered things" which of course is hilarious because today you can't possibly retain ALL the know…

>"in my days, we had books and we remembered things" which of course is hilarious it isn't hilarious, it's true. My father (now in his 60s) who came from a blue collar background with very little education taught himself programming by manually copying and editing software out of magazines, like a lot of people his age. I teach students now who have access to all the information in the world but a lot of them are qui…

Information 30 years ago was more difficult to obtain. It required manual labor but in todays' context there was not much information to be consumed. Today, we have the opposite - a huge vast of information that is easy to obtain but to process? Not so much. Decline is unavoidable. Human intelligence isn't increasing at the pace advancements are made.

Re: Using LLMs at Oxide

#154

Post by Madmallard »

adastra22 wrote: post #148
fastball wrote: post #124

Earlier quoted context omitted.

Has anything like this worked its way through the courts yet?

Yes, training is considered fair use, and output is non-copyrightable / public domain. With many asterix and footnotes, of course.

Don't see how output being public domain makes sense when they could be outputting copyrighted code.

Shouldn't the right's extend forward and simply require the LLM code to be deleted?

Re: Using LLMs at Oxide

#155

Post by yard2010 »

theresistor wrote: post #128
dboreham wrote: post #111

Earlier quoted context omitted.

Is there current generation LLMs do this? I suppose I mean "do this any more than human developers do".

A very recent example: https://github.com/ocaml/ocaml/pull/14369

>> Here's my question: why did the files that you submitted name Mark Shinwell as the author?

> Beats me. AI decided to do so and I didn't question it. I did ask AI to look at the OxCaml implementation in the beginning.

This shows that the problem with AI is philosophical, not practical

Re: Using LLMs at Oxide

#156

Post by Madmallard »

hexo wrote: post #138

"LLMs are amazingly good at writing code" that one was good. I cant stop laughing.

I wrote an entire multiplayer game in XNA that I've tried repeatedly to get LLMs to translate to javascript

it's just utterly hopeless how bad they are at doing it

even if I break it down into parts once you get into the stuff that actually matters i.e. the physics, event handling, and game logic induced by events, it just completely falls apart 100% of the time

Re: Using LLMs at Oxide

#157

Post by ares623 »

philippta wrote: post #145

> LLM-generated code should not be reviewed by others if the responsible engineer has not themselves reviewed it. To extend that: If the LLM is the author and the responsible engineer is the genuine first reviewer, do you need a second engineer at all? Typically in my experience one review is enough.

Yeesss this is what I’ve been (semi-sarcastically) thinking about. Historically it’s one author and one reviewer before code gets shipped.

Why introduce a second reviewer and reduce the rumoured velocity gained by LLMs? After all, “it doesn’t matter what wrote the code” right.

I say let her rip. Or as the kids say, code goes brrr.

Re: Using LLMs at Oxide

#158

Post by menaerus »

Madmallard wrote: post #154
adastra22 wrote: post #148

Earlier quoted context omitted.

Yes, training is considered fair use, and output is non-copyrightable / public domain. With many asterix and footnotes, of course.

Don't see how output being public domain makes sense when they could be outputting copyrighted code. Shouldn't the right's extend forward and simply require the LLM code to be deleted?

First, you have to prove it that it produced the copyrighted code. The question is what copyrighted code is in the first place? Literal copy-paste from source is easy but I think 99% of the time this isn't the case.

Re: Using LLMs at Oxide

#159

Post by Iridescent_ »

> Oxide employees bear responsibility for the artifacts we create, whatever automation we might employ to create them.

Yes, allow the use of LLMs, encourage your employees to use them to move faster by rewarding "performance" regardless of risks, but make sure to place responsibility of failure upon them so that when it happens, the company culture should not be blamed.

Re: Using LLMs at Oxide

#160

Post by tizzy »

The idea that LLMs are amazing at comprehension but we are expected to read original documents seems contradictory to me? I’m also wary of using them as editors and losing the writers voice as that feels heavily prompt dependent and whether or not the writer does a final pass without any LLM. Asking someone else to re-write is losing your voice if you don’t have an opinion on how the re-write turns out
Post reply on HN