Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
portaouflop
6 days ago
|
parent
|
context
|
favorite
| on:
ChatGPT Atlas
It’s called the waluigi problem and is also part of the reason why you can never fully “censor” an LLM; there is always some jailbreak possible
Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: