this post was submitted on 09 Apr 2025
1299 points (96.2% liked)
196
17484 readers
1107 users here now
Be sure to follow the rule before you head out.
Rule: You must post before you leave.
Other rules
Behavior rules:
- No bigotry (transphobia, racism, etc…)
- No genocide denial
- No support for authoritarian behaviour (incl. Tankies)
- No namecalling
- Accounts from lemmygrad.ml, threads.net, or hexbear.net are held to higher standards
- Other things seen as cleary bad
Posting rules:
- No AI generated content (DALL-E etc…)
- No advertisements
- No gore / violence
- Mutual aid posts are not allowed
NSFW: NSFW content is permitted but it must be tagged and have content warnings. Anything that doesn't adhere to this will be removed. Content warnings should be added like: [penis], [explicit description of sex]. Non-sexualized breasts of any gender are not considered inappropriate and therefore do not need to be blurred/tagged.
If you have any questions, feel free to contact us on our matrix channel or email.
Other 196's:
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
Had it write some simple shader yesterday cause I have no idea how those work. It told me about how to use the mix and step functions to optimize for GPUs, then promptly added some errors I had to find myself. Actually not that bad cause now after fixing it I do understand the code. Very educational.
This is my experience using it for electrical engineering and programming. It will give me 80% of the answer, and the remainder 20% is hidden errors. Turns out the best way to learn math from GPT is to ask it a question you know the answer (but not the process) to. Then, reverse engineer the process and determine what mistakes were made and why they impact the result.
Alternatively, just refer to existing materials in the textbook and online. Then you learn it right the first time.
thank you for that last sentence because I thought I was going crazy reading through these responses.
Some people retain information easier by doing than reading.
ok, I finally figured out my view on this I believe. I was worried I was being a grumpy old man who was just yelling at the AI (still probably am, but at least I can articulate why I feel this is a negative reply to my concerns)
It's not reproducible.
I personally don't believe asking an AI with a prompt then "troubleshooting" it is the best educational tool for the masses to be promoted to each-other. It works for some individuals, but as you can see the results will always vary with time.
There are so many promotional and awesome educational tools that emphasize the "doing" part instead of reading. You don't need to ask an AI prompt then try to fix all the horrible shit when there is always a statistically likely chance you will never be able to solve it and the AI gave you an impossible answer to fix.
I get some people do it, some people succeed, and some people are maybe so lonely that this interaction is actually preferable since it seems like some weird sort of collaboration. The reality is that the AI was trained unethically and has so many moral and ethical repercussions that just finding a decent educator or forum/discord to actually engage with is whole magnitudes better for society and your own mental processes.
In this context AI-gen codescript sounds like a fast track to a final exam before qualification
Shaders are black magic so understandable. However, they're worth learning precisely because they are black magic. Makes you feel incredibly powerful once you start understanding them.
I used it yesterday because I couldn't get mastodon's version of http signing working. It spat out a shell script which worked, which is more than my attempts did.
I hate it so damn much
Same. They don't even use the standard that was finalised, they are using an RFC draft that expired in 2022.