Hacker News .hn
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
pplante
on April 5, 2023
|
parent
|
context
|
favorite
| on:
Our Approach to AI Safety
Has anyone seen an experiment where an LLM talks to another LLM instance?
IanCal
on April 5, 2023
|
next
[–]
I've done this, it's very effective for some things. One LLM is told to come up with a plan, the other is told to critique it and push for concrete actions.
chaxor
on April 5, 2023
|
prev
[–]
We've been doing it for many many years. It's trivial to do yourself
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
