Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
catigula
2 days ago
|
parent
|
context
|
favorite
| on:
"time to GPT-2", down to 2.91 hours
The thing about dangerous AI is that AI will by definition not be dangerous precisely until it is.
This isn't a good reason to be reckless.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
This isn't a good reason to be reckless.