Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
Study: LLMs are prone to social identity biases similar to the way humans are
(
nyu.edu
)
1 point
by
giuliomagnifico
8 months ago
|
hide
|
past
|
favorite
|
1 comment
giuliomagnifico
8 months ago
[–]
But fortunately:
> AI biases can be reduced by carefully selecting the data used to train these systems
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
> AI biases can be reduced by carefully selecting the data used to train these systems