Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
siva7
63 days ago
|
parent
|
context
|
favorite
| on:
U.S. government will decide who gets to use GPT-5....
So OAI are you also silently dumbing down your models when you detect "inappropriate topics" like Anthropic did with Fable?
consensus1
63 days ago
[–]
Fable didn't silently dumb it down. It printed a warning that it has detected a possible inappropriate topic and you are being switched back to Opus. I hate it, but it isn't dishonest.
grim_io
63 days ago
|
parent
[–]
Unless it thought you are trying to distill it, then it would silently sabotage you.
j_maffe
62 days ago
|
root
|
parent
[–]
Didn't they announce they're stopping that system?
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
