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

I'd rather read the prompt

claytonwramsey.com

Post reply on HN
Print view View original

741–750 of 880 posts

Re: I'd rather read the prompt

#741

Post by necovek »

tptacek wrote: post #725
necovek wrote: post #522

Earlier quoted context omitted.

If you don't want to engage in an honest discussion, please refrain from making assumptions: nobody mentioned "merging crap". I stopped clearly at guiding an LLM to make a code change, which is at best, a pull request.

How is that better? As a team lead, what would you think of a team member who consistently generated "crap" pull requests? You see the same thing in every argument with LLM skeptics. 'The code is bad. You don't even know what the code is doing." This is obviously false . A professional reads the code they commit and push. A professional doesn't push code they know to be bad.

"Let's identify specific, clear areas for improvement, and if they are not able to improve, let's fire them": it's as simple as that.

Teaching a human with motivation, potential and desire to learn is both easier, and more rewarding (for most humans), than attempting to teach LLM to write good code every time — humans tend to value their personal experiences more, whereas LLM relies more on the training corpus. So when I've seen people massage LLM output to be decent or excellent, it took them more time than it would have taken for them to write it from scratch without an LLM.

Which makes LLMs mostly a curiosity, and not a productivity booster. Can it get there? I hope it can, because that would be amazing.

Re: I'd rather read the prompt

#742

Post by dcre »

necovek wrote: post #523
dcre wrote: post #407

Earlier quoted context omitted.

A couple of examples from influential open source developers: Adam Wathan (Tailwind) agreeing with Mitchell Hashimoto that LLMs are making them more productive. "in their area of expertise" is not obvious from this post alone, but I am pretty confident from the way they talk about it that they are not exclusively using LLMs on side projects where they're inexpert. https://x.com/adamwathan/status/1911845073286803923 A…

"Saying" vs "demonstrating": I am careful with the words I am using. It's not just nitpicking: maybe it's only their perception they are more productive? Or maybe they are. But saying it is not convincing enough for me.

That's a good point, though it's worth thinking about what would count as a demonstration. Pointing to a PR that an LLM helped with is not enough because you don't know how long it would have taken without the use of LLMs. The counterfactual never really exists. What is the hard evidence vim or emacs makes people more productive? Using git over svn? Using search engines? Over time, we expect people to use the things that make them more productive.

Re: I'd rather read the prompt

#743

Post by tptacek »

necovek wrote: post #741
tptacek wrote: post #725

Earlier quoted context omitted.

How is that better? As a team lead, what would you think of a team member who consistently generated "crap" pull requests? You see the same thing in every argument with LLM skeptics. 'The code is bad. You don't even know what the code is doing." This is obviously false . A professional reads the code they commit and push. A professional doesn't push code they know to be bad.

"Let's identify specific, clear areas for improvement, and if they are not able to improve, let's fire them": it's as simple as that. Teaching a human with motivation, potential and desire to learn is both easier , and more rewarding (for most humans), than attempting to teach LLM to write good code every time — humans tend to value their personal experiences more, whereas LLM relies more on the training corpus. So w…

None of this responds to what I just wrote. Can you engage with the question I asked directly? Thanks!

Re: I'd rather read the prompt

#744

Post by kergonath »

ziotom78 wrote: post #509
kergonath wrote: post #288

Earlier quoted context omitted.

> I'd love to hear any counterpoints from folks who have used LLMs lately to get academic or creative writing done They’re great at proofreading. They’re also good at writing conclusions and abstracts for articles, which is basically synthesising the results of the article and making it sexy (a task most scientists are hopelessly terrible at). With caveats: - all the information needs to be in the prompt, or they wil…

Point two is critical. I have found that the best way for me is to avoid using copy-and-paste. Instead, I put the browser on the right corner of the screen and my text editor on the left, then transcribe the text word by word by typing it using the keyboard. In this way, my natural laziness is less likely to accept words, expressions, and sentences that are perhaps okay-ish but not 100% following my taste.

Good process, I will try this as well.

Re: I'd rather read the prompt

#745

Post by klibertp »

alexvitkov wrote: post #715
klibertp wrote: post #639

Earlier quoted context omitted.

> It has value because a student can only get it by learning and problem-solving. No. That's how it should be, but the reality looks different: it has value because it shows that someone spent 3+ years doing what they were told to do, enduring all the absurdities they were subjected to in the meantime. Whatever means they used to cheat don't matter, since they still worked on what someone told them to and produced re…

This is a very bold statement, you should consider substantiating it at least a little, maybe give just one example on what those absurdities collee students have to endure are, that prove they're good corporate drones? From what I remember it was 4 years of learning stuff I signed up to learn, occasionally being quizzed on said stuff, and then they gave me a paper that claims I know the stuff.

You might live in another country where the situation is different, or maybe my experiences are outdated (20 years later, they might be). What I remember is unfairness and corruption, feudal-like relationships, tons of wasted time on things I didn't sign up for and things I already knew but couldn't say I know for fear of retaliation. I was still better than people going to private institutions: on top of all that, they were also extorted for money at every turn. ...thinking about it, yeah, I really hope my experience is outdated.

Re: I'd rather read the prompt

#746

Post by dyauspitr »

lazide wrote: post #560
dyauspitr wrote: post #541

Earlier quoted context omitted.

The problem is there is no reliable way to test for LLM usage unlike for doping. There’s no way to police this problem.

Sure there is - evaluate the candidates actual skill level without the ability to use automated tools. The current and old school way is a proctored exam.

Testing doesn’t work for everything, especially for dissertations, publications, research etc. That being said I’m wholeheartedly in favor of using extensive in class testing when applicable.

Re: I'd rather read the prompt

#747

Post by qoez »

sn9 wrote: post #221

> I should hope that the purpose of a class writing exercise is not to create an artifact of text but force the student to think; a language model produces the former, not the latter. It's been incredibly blackpilling seeing how many intelligent professionals and academics don't understand this, especially in education and academia. They see work as the mere production of output, without ever thinking about how that…

To play devils advocate: There's so many times in school where I remember having to do BS tasks which could have been better spent working on side projects instead. If the students are smart that's how they'll leverage these tools. The vast majority won't but some will.

Re: I'd rather read the prompt

#748

Post by biker142541 »

dyauspitr wrote: post #541
sgustard wrote: post #413

Earlier quoted context omitted.

Why do some countries cheat in the Olympics? Because it is no longer a contest of human achievement, it's just about the medals as a symbol of national glory. Of course: once all countries are doping, the medals will become meaningless. College degrees will suffer the same fate if everyone cheats to get them.

The problem is there is no reliable way to test for LLM usage unlike for doping. There’s no way to police this problem.

Diversity in evaluation methodology is a reliable way for many subject. Even short, minimal handwritten exams would help to assess student understanding at a few key checkpoints and should should be reintroduced.

Re: I'd rather read the prompt

#749

Post by bost-ty »

kouru225 wrote: post #341

Ever since AI came out I’ve been talking about the prompt to output ratio. We naturally assume that the prompt will be smaller than the output just because of the particulars of the systems we use, but as you get more and more particular of what you want, the prompt grows while the output stays the same size. This is logical. If instead of writing an essay, I just describe what I want the essay to say, the descriptio…

> In that way, the prompt is more interesting, and I can’t tell you how many times I’ve gone to go write a prompt because I dunno how to write what I wanna say, and then suddenly writing the prompt makes that shit clear to me.

Bingo. It can be a rubber duck that echoes your mistakes back. Unfortunately, as other commenters have pointed out, the prompt may not be as interesting/iterative as we might suppose: "Here's the assignment, what's the answer".

Post reply on HN