this post was submitted on 21 May 2025
998 points (97.8% liked)
Technology
71396 readers
3406 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related news or articles.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
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
Computers with some encyclopedia, but no GPTs are fine, no?
If a kid can write and train a mini-GPT trainable on that encyclopedia, then maybe they deserve the mark for desperation and ingenuity and being a fucking new Leonardo.
Open book and calculators would seem reasonable. No communication or searching devices.
No communication - of course, but about search - I don't think having a Wikipedia snapshot with search is bad.
GPTs are fine, if you learn to disrespect their output and fix it before presenting it as your own.
Actually, taught that way, GPT may be a tool for teaching critical thinking - if the professors aren't too lazy to mark down the garbage output.
A lot of people "have trouble getting started" - in all kinds of endeavors. Once you get them rolling, they can see the pattern and do it for themselves next time. If the AI glop gets lucky and copied decent argument from beginning to end (something I've seen it fail spectacularly at many times), then that can help jumpstart people who are stuck, but only if they can recognize when it's just a bunch of glop.
Really, if would be better for them to read a bunch of samples for themselves (which is what the AI does) and hopefully they can get the pattern. What I think is a horrible approach is to sit in a lecture hall and listen to a little guy down front drone in a monotone about the theory of what you are supposed to do, then try to synthesize from the fragments of what you understood from that what is expected. Samples to work from are much more efficient.