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

41–50 of 880 posts

Re: I'd rather read the prompt

#41

Post by tptacek »

I have a lot of sympathy for the author's position but I may have missed the point in the article where he explained why clarity of writing and genuineness of human expression was so vital to a robotics class. It's one thing for an instructor to appreciate those things; another for them to confound their own didactic purpose with them. This point seems obvious enough that I feel like I must have missed something.

As always, I reject wholeheartedly what this skeptical article has to say about LLMs and programming. It takes the (common) perspective of "vibe coders", people who literally don't care what code says as long as something that runs comes out the other side. But smart, professional programmers use LLMs in different ways; in particular, they review and demand alterations to the output, the same way you would doing code review on a team.

Re: I'd rather read the prompt

#42

Post by oncallthrow »

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.

There is no way to design such a system that is not cheatable. At the very least, someone could simply type out text from another window or device. On any normal operating system or browser, the user will be able to bypass whatever mechanism you have in place anyway.

Re: I'd rather read the prompt

#43

Post by hoppp »

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.

You can still just type the Ai response. Often when I generate larger code I type it instead of copy paste, that helps me understand it and spot issues faster

Re: I'd rather read the prompt

#44

Post by bonoboTP »

matthewdgreen wrote: post #36
robertlagrant wrote: post #18

Time to go back to writing essays in exams, live, on paper.

It's challenging. Assignments (and particularly programming assignments) were by far the larger and more difficult part of my CS degree, and also the place where I learned the most. I cannot imagine losing that portion of my education and just replacing it with a few exams.

That's how it works in Germany. Usually assignments are either optional or you just have to get 50% of the total assignment marks over the semester to be admitted to the exam (written or often oral, in person). Then your grade is entirely based on the exam. Hand-holding throughout the semester assignment-to-assignment, checking attendance etc. is more an Anglo-specific thing where students are treated as kids instead of adults.

Re: I'd rather read the prompt

#45

Post by jfengel »

Animats wrote: post #28

That's because the instructor is asking questions that merely require the student to regurgitate the instructor's text. To actually teach this, you do something like this: "Here's a little dummy robot arm made out of Tinkertoys. There are three angular joints, a rotating base, a shoulder, and an elbow. Each one has a protractor so you can see the angle. 1. Figure out where the end of the arm will be based on those th…

A LLM can't do that? I'm a little surprised.

(I know jack all about robotics but that sounds like a pretty common assignment, the kind an LLM would regurgitate someone else's homework.)

Re: I'd rather read the prompt

#46

Post by oncallthrow »

cortesoft wrote: post #5

> I say this because I believe that your original thoughts are far more interesting, meaningful, and valuable than whatever a large language model can transform them into. Really? The example used was for a school test. Is there really much original thought in the answer? Do you really want to read the students original thought? I think the answer is no in this case. The point of the test is to assess whether the stu…

> Do you really want to read the students original thought?

Why else would you become a teacher, if you didn't care about what your students think?

Re: I'd rather read the prompt

#47

Post by echelon »

bost-ty wrote: post #4

I like the author's take: it isn't a value judgement on the individual using ChatGPT (or Gemini or whichever LLM you like this week), it's that the thought that went into making the prompt is, inevitably, more interesting/original/human than the output the LLM generates afterwards. In my experiments with LLMs for writing code, I find that the code is objectively garbage if my prompt is garbage. If I don't know what I…

> I'd love to hear any counterpoints from folks who have used LLMs lately to get academic or creative writing done

I commented in another thread. We're using image and video diffusion models for creative:

https://www.youtube.com/watch?v=H4NFXGMuwpY

Still not a fan of LLMs.

Re: I'd rather read the prompt

#48

Post by ineedasername »

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…

Yes, depending on the model being used, endless text of this flavor isn't all that compelling to read:

"Tall man, armor that is robotic and mechanical in appearance, NFL logo on chest, blue legs".,

And so on, embedded in node wiring diagrams to fiddly configs and specialized models for bespoke purposes, "camera" movements, etc.

Re: I'd rather read the prompt

#49

Post by topkai22 »

sillysaurusx wrote: post #22

I think people who don’t like writing shouldn’t be forced to write, just like people who don’t like music shouldn’t be forced to play music. Ditto for math. Forcing people to do these things supposedly results in a better, more competitive society. But does it really? Would you rather have someone on your team who did math because it let them solve problems efficiently, or did math because it’s the trick to get the r…

Many of the things we teach in school aren’t just for the direct knowledge or skill. We largely don’t need to do arithmetic any more, but gaining the skill at doing it really improves our ability to deal with symbolic manipulation and abstraction.

I remember another parent ranting about their 3rd grade kids “stupid homework” since it had kids learning different ways of summing numbers. I took a look at the homework and replied “wow, the basics out set theory are in here!” We then had a productive discussion of how that arithmetic exercise led to higher math and ways of framing problems.

Similarly, writing produces a different form of thought than oral communication does.

History is a bit different, but a goal of history and literature is (or it least should be) to socialize students and give them a common frame of reference in society.

Finally there is the “you don’t know when you’ll need it defense.” I have a friend who spent most of the last 20 years as a roofer, but his body is starting to hurt. He’s pivoting to CAD drafting and he’s brushing off a some of those math skills he hated learning in school. And now arguing with his son about why it’s important.

Those are the fundamental defenses- that we are seeking not skills but ways of viewing the world + you don’t know what you’ll need. There are obviously limits and tradeoffs to be made, but to some degree yes, we should be forcing students (who are generally children or at least inexperienced in a domain) to things they don’t like now for benefits later.

Re: I'd rather read the prompt

#50

Post by necovek »

sillysaurusx wrote: post #22

I think people who don’t like writing shouldn’t be forced to write, just like people who don’t like music shouldn’t be forced to play music. Ditto for math. Forcing people to do these things supposedly results in a better, more competitive society. But does it really? Would you rather have someone on your team who did math because it let them solve problems efficiently, or did math because it’s the trick to get the r…

"Forcing" is a bit strong IMHO — I believe we've instead lost track of what is "passable", and everyone in higher education should be able to reach that and score a passing grade (D? C?).

Maybe professors are too stringent with their evaluation, or maybe they are not good at teaching people what a passable writing style is, or maybe students simply don't want to accept that if they don't excel at writing, a D or a C is perfectly fine. Perhaps teachers that look for good writing should have separate tests which evaluate students in both scenarios: with and without LLM help.

The same holds true for math: not everybody needs to know how to deduce a proof for every theorem, but in technical sciences, showing that ability and capability will demonstrate how much they are able to think and operate with precision on abstract concepts, very much like in programming. Even if coursework is a bit repetitive, practice does turn shallow knowledge into operational knowledge.

Post reply on HN