Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
jagged-chisel
6 days ago
|
parent
|
context
|
favorite
| on:
GPT-5 outperforms federal judges in legal reasonin...
Presumably all these agents would have been trained on different data, with different viewpoints? Otherwise, what makes them different enough from each other that such a "conversation" would matter?
help
qwertox
6 days ago
[–]
Different skills or plugins, different views and different tools for the analysis of the same object. Then the debate starts.
reply
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: