Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
zamadatix
5 months ago
|
parent
|
context
|
favorite
| on:
GPT-5
"+100 points" sounds like a lot until you do the ELO math and see that means 1 out of 3 people still preferred Claud Opus 4's response. Remember 1 out of 2 would place the models dead even.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: