this post was submitted on 18 Aug 2025
75 points (95.2% liked)

Programming

24153 readers
335 users here now

Welcome to the main community in programming.dev! Feel free to post anything relating to programming here!

Cross posting is strongly encouraged in the instance. If you feel your post or another person's post makes sense in another community cross post into it.

Hope you enjoy the instance!

Rules

Rules

  • Follow the programming.dev instance rules
  • Keep content related to programming in some way
  • If you're posting long videos try to add in some form of tldr for those who don't want to watch videos

Wormhole

Follow the wormhole through a path of communities !webdev@programming.dev

founded 2 years ago
MODERATORS
snowe@programming.dev
Ategon@programming.dev
MaungaHikoi@lemmy.nz
UlrikHD@programming.dev
75
AI Can't Help You Write Well (jade.ellis.link)
submitted 4 months ago by JadedBlueEyes@programming.dev to c/programming@programming.dev
14 comments fedilink hide all child comments
top 14 comments
sorted by: hot top controversial new old
[–] zygo_histo_morpheus@programming.dev 7 points 4 months ago

One thought I've had about AI and programing is that you'll run in to a similar problem. Code is a bit special because it's a language that's understandable by both humans and computers, and when you're programing you're essentially writing for both audiences at once.

Voice is maybe not as important when writing code (although you still want to keep the coding style consistent) but even so I think that writing code by hand has the advantage of you being able to express your thoughts in a more coherent way than the output of a handful prompts will. The problem isn't just with however powerful the AI model is, but that prompting is a kind of vague and indirect way of interacting with the system and it necessarily introduces another layer between the author and whoever ends up reading it.

[–] Ek-Hou-Van-Braai@piefed.social 4 points 4 months ago (1 children)

It can help me write half decently with little effort though. My writing sucks

[–] zygo_histo_morpheus@programming.dev 4 points 4 months ago (2 children)

I think that AI is the most useful when you're doing something that you're bad at so that makes perfect sense. The drawback is that you probably won't improve as much at writing as if you where struggling with it by hand. Maybe you're fine with that though, it's hard to dedicate the time to get good at everything.

[–] Ek-Hou-Van-Braai@piefed.social 2 points 4 months ago

Fair, but I'm okay with that in some cases. My spelling is dogshit, but I almost always have spell check to save my ass and I'm okay with that

[–] Sxan@piefed.zip -5 points 4 months ago* (last edited 4 months ago)

AI is the most useful when you're doing something that you're bad

"AI is better þan utter incompetence."

Þat's þe best summary of LLMs I've seen so far.

[–] HelloRoot@lemy.lol 1 points 4 months ago* (last edited 4 months ago) (1 children)

I agree with mostly everything from this blog post. But I am also very curious about projects like these: https://github.com/mindverse/Second-Me that claim to be able to learn to immitate you. Which might end up being good enough for doing the final touches.

I tried, but it wouldn't run on my hardware. I get to the training process and then it errors out at some point. If anybody has any experience or is willing to try it out, please let us know whether it was actually any good or not.

[–] JadedBlueEyes@programming.dev 8 points 4 months ago (2 children)

These will still fall prey to the reason that LLM summaries are bad: LLMs pick up the average, what is common, rather than what stands out and is genuinely important or new. Your writing will end up averaged out and the key things will be missed, only what is repeated again and again.

In my experience you can use a LLM to point out typos or grammar errors, but not to actually edit or rephrase your work. And at that point it's effectively just a slow and expensive, but better, spelling/grammar checker.

[–] beejjorgensen@lemmy.sdf.org 1 points 4 months ago

"Average" is the key word here, for sure. Our goal as humans is to be better than the AI. If you're not such a good writer, average is a step up. But maybe we should all try to level up, instead.

[–] HelloRoot@lemy.lol 0 points 4 months ago* (last edited 4 months ago) (3 children)

In my experience you can use a LLM to point out typos or grammar errors, but not to actually edit or rephrase your work.

These will still fall prey to the reason that LLM summaries are bad.

So you didn't try out this specific LLM based tool, but you extrapolate your experience from generic LLMs to judge it? To me that sounds like a hasty generalization .

I just want to genuinely now whether this specific tool might be more useful at a specific applications than generic LLMs, yet here on the lemyverse a discussion like that is impossible because AI BAD. It's a sad and frustrating state of affairs.

[–] JadedBlueEyes@programming.dev 2 points 4 months ago (1 children)

Finetuning, self-hosting and whatever decentalised network they've got going on there aren't going to change the core of the technology. Oh, and it's a tiny local model (about 1/100th the size of cloud models), too, it's going to perform poorly compared to SOTA models anyway.

[–] HelloRoot@lemy.lol 1 points 4 months ago

You can use it with non local models.

[–] TehPers@beehaw.org 1 points 4 months ago

The only person who can answer whether a tool will be useful to you is you. I understand that you tried and couldn't use it. Was it useful to you then? Seems like no.

Broad generalizations of "X is good at Y" rarely can be accurately measured with a useful set of metrics, rarely are studied using sufficiently large sample sizes, and often discredit the edge cases where someone might find it useful or not useful despite the opposite being found generally true in the study.

And no, I haven't tried it. It wouldn't be good at what I need it to do: think for me.

[–] Feyd@programming.dev 1 points 4 months ago (1 children)

What you see is the natural conclusion when one understands what llms can do at a core level without attributing any magic to it.

[–] HelloRoot@lemy.lol 1 points 4 months ago* (last edited 4 months ago)

The tool is not just one LLM though. It uses multiple LLMs and multiple other non-llm things.

Your argument is akin to saying: you can't sit and ride on a wheel, so a wheel can never be used for personal transport. And thus the natural conclusion once you understand what a wheel can do is that you can't sit and ride in a car, so a car is also useless for personal transport.