this post was submitted on 20 Jul 2023
69 points (93.7% liked)

ChatGPT

1479 readers
3 users here now

A community to discuss about ChatGPT and AI. Not affiliated with OpenAI.

Rules:

  1. NSFW must be marked as such

  2. No porn

  3. Don't be toxic

  4. jailbreaks are allowed for now at least

  5. No trashposts.

  6. Keep the memes to a minimum

founded 1 year ago
MODERATORS
GoldenDoge@lemmy.ml
SeeleLowe@lemmy.ml
Senseibull@lemmy.ml
69
Over just a few months, ChatGPT went from correctly answering a simple math problem 98% of the time to just 2%, study finds (fortune.com)
submitted 1 year ago by tracyspcy@lemmy.ml to c/chatgpt@lemmy.ml
8 comments fedilink hide all child comments
you are viewing a single comment's thread
view the rest of the comments
[–] taladar@sh.itjust.works 19 points 1 year ago (3 children)

A system that has no idea if what it is saying is true or false or what true or false even mean is not very consistent in answering things truthfully?

[–] tracyspcy@lemmy.ml 4 points 1 year ago* (last edited 1 year ago)

Wait for the next version which will be trained on data that includes gpt generated word salad

[–] VM_Abrantes@kbin.social 1 points 1 year ago

Detroit: Become Human moment.

[–] intensely_human@lemm.ee -1 points 1 year ago

No that is not the thesis of this story. If I’m reading the headline correctly, the rate of its being correct has changed from one stable distribution to another one.