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

291–300 of 880 posts

Re: I'd rather read the prompt

#291

Post by bluefirebrand »

deadbabe wrote: post #245
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…

Someday, LLMs are going to cost actual serious money to use and people will have “LLM bills” to pay every month just to have some semblance of a brain at work. Maybe a portion of salary can just be deducted directly from a paycheck to pay for LLM tools.

Companies will have to provide the LLM just like retail stores have to provide cash registers for cashiers to use

Re: I'd rather read the prompt

#293

Post by palata »

imhoguy wrote: post #159
laurentlb wrote: post #127

There are many ways to use LLMs. The issue, IMO, is that some people throw in a one-shot, short prompt, and get a generic, boring output. "Garbage in, generic out." Here's how I actually use LLMs: - To dump my thoughts and get help organizing them. - To get feedback on phrasing and transitions (I'm not a native speaker). - To improve tone, style (while trying to keep it personal!), or just to simplify messy sentences…

Exactly it is a tool which needs skill to use. I would add extra use of mine: - To "Translate to language XYZ", and that is not sometimes strightforward and needs iterating like "Translate to language used by living in " and so on. And the author is right, I use it as 2nd-language user, thus LLM produces better text than myself. However I am not going to share the prompt as it is useless (foreign language) and too me…

For what it's worth, I think that sending a message translated to a foreign language you don't master is the worst thing you can do.

You speak English? Write and send your message in English. The receiver can copy-paste it in a translator. This way, they will know that they are not reading the original. So if your translated message sounds inaccurate, offensive or anything like that, they can go back to your original message.

Re: I'd rather read the prompt

#294

Post by palata »

echelon wrote: post #35
Ancalagon wrote: post #8

I fully support the author’s point but it’s hard to argue with the economics and hurdles around obtaining degrees. Most people do view obtaining a degree as just a hurdle to getting a decent job, that’s just the economics of it. And unfortunately the employers these days are encouraging this kind of copy/paste work. Look at how Meta and Google claim the majority of the new code written there is AI created? The world…

> I fully support the author’s point I don't. I think the world is falling into two camps with these tools and models. > I now circle back to my main point: I have never seen any form of create generative model output (be that image, text, audio, or video) which I would rather see than the original prompt. The resulting output has less substance than the prompt and lacks any human vision in its creation. The whole po…

> The latter is absolutely a great use case for AI.

The video is not exactly great, IMO.

Re: I'd rather read the prompt

#295

Post by recursivecaveat »

scarface_74 wrote: post #10

I don’t know anything about the subject area, so I don’t know if this captures enough to get a good grade. But I’m curious if anyone could tell whether the last answer were AI generated if I copied and pasted. These are the iterations I go through when writing long requirement documents/assessments/statements of work (consulting). Yes I know the subject area for which I write assessments and know if what is generated…

I mean this is not blind obviously, but it feels unnaturally enthusiastic/conversational to me. Maybe for like an video script it would fit, but for a requirements document or something it is a little oddly 'sauced up', as if someone put an extra pass through it to try to make it entertaining to read.

Re: I'd rather read the prompt

#296

Post by ozim »

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…

I always get triggered when people argue against „rote memorization” - but it also is technique that builds up knowledge, skills and experience.

Even if one won’t need that specific know how after exams - just realization how much one can memorize and trying out some approaches to optimize it is where people grow/learn.

Re: I'd rather read the prompt

#297

Post by palata »

theturtletalks wrote: post #274
palata wrote: post #256

Earlier quoted context omitted.

I would argue that if you are losing a consequent amount of your time fixing grammar, then it sounds like you need to spend that time to improve your grammar skills. > We need to teach students to use AI for deeper thinking by asking better questions. Same thing here: the whole point of learning critical thinking is that you don't need to ask someone/something else. Teaching you how to ask the LLM to do it for you is…

You’re getting to the crux of the argument, knowing when to use AI. Doing or learning “It” in 2025 means using AI whether to understand it better or use it to get better grades. My take is teach them to get better at asking questions and then teach them when to use their own understanding to change their answer for the better. How many times has an AI’s answer been 5/10 and with a few fixes it’s a 9/10. That comes wi…

> Doing or learning “It” in 2025 means using AI whether to understand it better or use it to get better grades.

Depends, I think. If we are talking about writing an essay (and I believe we are), then the LLM is somewhere between useless and counter-productive.

Of course, if the LLM is used to understand an RFC (I would debate how useful it is for that, but that's another discussion), then it's different. The goal was to understand the RFC, it doesn't really matter how you did it. But the goal of writing an essay is not to end up with a written essay at all. Nobody cares about it, you can burn it right after it's graded.

Re: I'd rather read the prompt

#298

Post by ashwinsundar »

xmorse wrote: post #20

I am thinking about creating a proof-of-writing signature. Basically an editor with an "anti-cheat", you can't paste text into it. It signs your text with a public key.

For everyone pointing out that this idea can be cheated by just typing AI-generated text into the editor - add an AI-detector to the editor. Gamify the whole thing by making a leaderboard of people with the lowest AI-detector-similarity score across things that they have "written"

In a class setting, maybe make the AI-detection an element of take-home assignments - whoever gets the lowest AI-similarity score gets a few points of extra credit or something

As for computer science courses, I'm guessing it's hard to not write simple code that appears AI-generated...so maybe that kind of work needs a written summary to go along with the code as well

Re: I'd rather read the prompt

#299

Post by easygenes »

fallinditch wrote: post #289
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…

It's been obvious since ChatGPT blew up in early 2023 that educators had to rethink how they educate. I agree that this situation that the author outlines is unsatisfactory but it's mostly the fault of the education system (and by extension the post author). With a class writing exercise like the author describes, of course the students are going to use an LLM, they would be stupid not to if their classmates are usin…

Amusingly, when I asked o3 to propose changes to the education system which address the author's complaints wrt writing assignments, one of the first things it suggested was transparent prompt logging (basically what the author proposes).

https://chatgpt.com/share/6817fe76-973c-8011-acf3-ef3138c144...

Re: I'd rather read the prompt

#300

Post by zeroq »

palata wrote: post #262
zeroq wrote: post #255

Earlier quoted context omitted.

You're missing trees for forest. When I was a kid and got an assignement for writing an essey about "why good forces prevailed in Lords of the Rings" as a gate check to see if I actually read the novel I had three choices: (a) read the novel and write the essey myself (b) find an already written essey - not an easy task in pre-internet era but we had books with esseys on most common topics you could "copy-paste" - an…

essay* (I don't usually do that, but it appears so many times in the first few sentences that I had to do it here) I agree with your points, though, but I think that they are in agreement with the comment you are answering to...

Hehe, it's fine, at least it proves that the post was written by human. ;)

And yeah, and revisiting the OP we're on the same track.

Post reply on HN