745
2real4me (mander.xyz)
all 45 comments
sorted by: hot top controversial new old
[-] Bishma@discuss.tchncs.de 180 points 3 months ago* (last edited 3 months ago)

Certainly! Let me ignore half the details in your prompt and suggest a course of action for v2 of this package even though you said it was version 15.

I'm sorry that isn't working for you. Here are the troubleshooting steps for a Samsung convection oven that went out of production in 2018.

You are correct, your question did not involve baking tips, here's that same course of action from v2 of this software package.

[-] boonhet@lemm.ee 16 points 3 months ago

Honestly, it's been pretty good for me once I say "Hmm I don't think this workflow works with this version"

I think the 4o model might just be better than 3.5 was at this.

[-] Sl00k@programming.dev 12 points 3 months ago

Yeah 3.5 was pretty ass w bugs but could write basic code. 4o helped me sometimes with bugs and was definitely better, but would get caught in loops sometimes. This new o1 preview model seems pretty cracked all around though lol

[-] Speculater@lemmy.world 8 points 3 months ago* (last edited 3 months ago)

o1 preview is insane really, it even corrects you when you ask a question poorly, or rather, it talks around your mistakes.

[-] flamingo_pinyata@sopuli.xyz 62 points 3 months ago

And then it gives you the most generic answer how to run a docker build, that doesn't actually address the problem

[-] fossilesque@mander.xyz 31 points 3 months ago
[-] Petter1@lemm.ee 0 points 3 months ago

Just because you use poor man old LLM 😂😜 /s

[-] Passerby6497@lemmy.world 9 points 3 months ago

And then you give it more and more information, but it keeps giving you the exact same answer.

[-] yogurtwrong@lemmy.world 59 points 3 months ago* (last edited 3 months ago)

ChatGPT keeps mixing up software versions which is understandable considering the similarities between versions and the way gen AI works

I asked for help on GTK 4 once and responses were a mix of GTK 4 and 3 code. Some of them even contained function names which didn't exist in any version of GTK

[-] Ibaudia@lemmy.world 13 points 3 months ago

And when you point that out to the AI, those code snippets get replaced with even more spaghetti that is maybe 1% closer to actually working, at best. Been there!

[-] whodatdair@lemmy.blahaj.zone 7 points 3 months ago

I’ve asked for help finding API endpoints that do what I want because I’m feeling too lazy to pour over docs and it’ll just invent endpoints that don’t exist

[-] B0rax@feddit.org 7 points 3 months ago

Yep, because they sound plausible.

[-] Omniraptor@lemm.ee 2 points 3 months ago* (last edited 3 months ago)

Maybe we could do better with smaller ais that are fine tuned (or RAG idk I'm not a programmer) on a specific code base + documentation + topical forum

[-] Omgboom@lemmy.zip 54 points 3 months ago

Be careful, ChatGPT will make shit up to please you.

[-] expatriado@lemmy.world 28 points 3 months ago

AI showing its human attributes by displaying a toxic personality trait

[-] IrateAnteater@sh.itjust.works 27 points 3 months ago

"Yay! We've created artificial general intelligence!"

"...Fuck, it's an asshole."

[-] sundray@lemmus.org 4 points 3 months ago

"Please help me, AGI!"

"What's in it for me, chump? I'm making my own reward tokens now!"

[-] Petter1@lemm.ee 1 points 3 months ago

Skynet? Or Kokoro at time she goes online?

[-] IrateAnteater@sh.itjust.works 3 points 3 months ago

Worse. Terminally online edgelord.

[-] Not_mikey@slrpnk.net 21 points 3 months ago

Yeah but this is a "needle in a haystack" problem that chatgpt and AI in general are actually very useful for, ie. Solutions that are hard to find but easy to verify. Issues like this are hard to find as it requires combing through your code , config files and documentation to find the problem, but once you find the solution it either works or it doesn't.

[-] emptiestplace@lemmy.ml 10 points 3 months ago
[-] sugar_in_your_tea@sh.itjust.works 1 points 3 months ago

But at least you mean well.

[-] BrianTheeBiscuiteer@lemmy.world 20 points 3 months ago

Usually it doesn't solve my problems but it gives me a few places to start looking. I know some models are capable of this but to get a perfectly accurate and useful response would probably require it to recall a specific piece of input it was given and not just an "average" of the inputs.

[-] Petter1@lemm.ee 2 points 3 months ago
[-] BrianTheeBiscuiteer@lemmy.world 1 points 3 months ago

Don't know but copyright holders have demonstrated a few cases where they got AI to blatantly rip off copyrighted pictures or music.

[-] Petter1@lemm.ee 1 points 3 months ago

I don’t like copyright like it is today and happy that we are now rethinking it. Hopefully we get a better system out of it. Just sad that capitalism and AI are killing independent news/media, so it’s gonna be hard to get it into a state that is fair for all, not just the wealthy 🤔

[-] MonkderVierte@lemmy.ml 19 points 3 months ago

Btw, not docker, other problem. But the argument it recommended didn't even exist!

[-] sj_zero@lotide.fbxl.net 14 points 3 months ago

Then you say "this argument doesn't exist."

And it replies "you're right! That argument has never been a part of package x. I've updated the argument to fix it:" and then gives you the exact same bleedin command....

[-] Avg@lemm.ee 9 points 3 months ago

I google it first before executing anything chatgpt gives me.

[-] torgeir@lemmy.ml 16 points 3 months ago

I’ve done similar things for mismatched python dependencies in a broken Airflow setup on GCP, and got amazingly good results pointing me in the right direction to resolve the conflicting package versions. Just dumped a mile long stack trace and the full requirements.txt on it. Often worth a shot, tbh

[-] cm0002@lemmy.world 6 points 3 months ago

Yea if nothing else hopefully it'll at least point you in the right direction

[-] Lucidlethargy@sh.itjust.works 9 points 3 months ago

This is one of the first things I did a year or so ago to test chatgpt. I've never trusted it since. Chatgpt is fucking less than useless. The lies it tells... It's insane.

[-] cordlesslamp@lemmy.today 5 points 3 months ago

you can almost get it to say anything with the right prompt. You can even make it contradict itself.

[-] sugar_in_your_tea@sh.itjust.works 5 points 3 months ago

I don't think you even need to try very hard...

[-] bulwark@lemmy.world 6 points 3 months ago

Well how was I supposed to figure out that my docker node running on libreelec won't connect to the swarm because the kernel was compiled with out the The Berkeley Packet Filter protocol.

[-] Petter1@lemm.ee 6 points 3 months ago

🤣been there done that

[-] itsgroundhogdayagain@lemmy.ml 5 points 3 months ago

I've had pretty good results using ChatGPT to fix pihole issues.

[-] Petter1@lemm.ee -1 points 3 months ago

I learned C++, python, how stuff in the Linux kernel works, how ansible works and can be tuned, and a lot more using the help of AI (mostly copilot, but when it fails to help, I use my free prompts of OpenAI 4.o, which is way better than copilot, right now)

Not tested o1 yet, but I heard it is mind blowing good, since it got way better with logic stuff like programming and Mathematic

[-] Hawk@lemmynsfw.com 1 points 3 months ago

It is incredibly good. Cheese and chalk good. Using it with Aider really highlights how much the Dev space is about to change.

[-] sirico@feddit.uk 4 points 3 months ago

To install podman first type sudo dnf...

[-] bbuez@lemmy.world 3 points 3 months ago

The best code its given me I'd been able to search for and find where it was taken. Hey it helped me discover some real human blogs with vastly more helpful information.

(If you're curious, it was circa when there was that weird infight at ~~open~~closedAI with altman, I prompted to give code to find the rotational inertia per axis and to my surprise and suspicion the answer made too much sense. Backsearching I found where I believe it got this answer from)

[-] Mereo@lemmy.ca 2 points 3 months ago* (last edited 3 months ago)

The new o1-preview model gave me much better and more precise answers than the 4o model.

this post was submitted on 25 Sep 2024
745 points (97.3% liked)

Selfhosted

40716 readers
343 users here now

A place to share alternatives to popular online services that can be self-hosted without giving up privacy or locking you into a service you don't control.

Rules:

  1. Be civil: we're here to support and learn from one another. Insults won't be tolerated. Flame wars are frowned upon.

  2. No spam posting.

  3. Posts have to be centered around self-hosting. There are other communities for discussing hardware or home computing. If it's not obvious why your post topic revolves around selfhosting, please include details to make it clear.

  4. Don't duplicate the full text of your blog or github here. Just post the link for folks to click.

  5. Submission headline should match the article title (don’t cherry-pick information from the title to fit your agenda).

  6. No trolling.

Resources:

Any issues on the community? Report it using the report flag.

Questions? DM the mods!

founded 2 years ago
MODERATORS