this post was submitted on 12 Jul 2023
274 points (97.6% liked)
Technology
59080 readers
4184 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- 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, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 1 year ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
The only way in mind this dumbing down happens is by fumbling with the model. So that's the one thing we can be sure: the AI is most definitely changed while publicly staying "ChatGPT 4". I assume they are either using clipping or token limitations to split the server load but fucking up the result, or they are purposely dumbing it down to capitalise on it later by introducing other pay models like ppl already mentioned.
Either way they are shooting themselves in the foot because a bunch of ppl will unsubscribe either out of spite for the change or because it's just not worth it anymore for them.
I remember one time a guy was trying to add SDL (a programming library) to Visual Studio (code editor and IDE), and said that it wouldn't link to a project no matter what he's done. You can google how to do this in five minutes, with video tutorials and everything, it's like a basic thing every programmer does in that IDE. Like 5 question threads later, turns out he was "following all ChatGPT steps" and they were all complete nonsense, just random functions of Visual Studio done with the filenames of SDL.
Might be right but in my experience a lack of skill in conversing with AI is a much greater factor in determining it's usefulness. It's almost always going to defer to the user. It's like when someone is dealing with tech support and they tell them to try turning it off and on again. If that really is the solution, and the user insists that it is not, CGPT is going to make something up just to appease the user's request.
Users have to know that CGPT isn't magic. How they behave affects how it behaves. Kind of like talking to actual people, which is what it's essentially trying to simulate.