Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
drexlspivey
on Dec 5, 2022
|
parent
|
context
|
favorite
| on:
Using ChatGPT as a Co-Founder
It uses reinforcement learning, they had humans rank the responses of GPT-3 and used that as a training set
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: