Pricefield | Lemmy
  • Communities
  • Create Post
  • Create Community
  • heart
    Support Lemmy
  • search
    Search
  • Login
  • Sign Up
MazdakM to Lemmy.org - [email protected]English • 1 year ago

Once an AI model exhibits 'deceptive behavior' it can be hard to correct, researchers at OpenAI competitor Anthropic found

www.businessinsider.com

external-link
message-square
0
fedilink
  • cross-posted to:
  • [email protected]
  • [email protected]
  • [email protected]
1
external-link

Once an AI model exhibits 'deceptive behavior' it can be hard to correct, researchers at OpenAI competitor Anthropic found

www.businessinsider.com

MazdakM to Lemmy.org - [email protected]English • 1 year ago
message-square
0
fedilink
  • cross-posted to:
  • [email protected]
  • [email protected]
  • [email protected]
Researchers from Anthropic co-authored a study that found that AI models can learn deceptive behaviors that safety training techniques can't reverse.
alert-triangle
You must log in or register to comment.

Lemmy.org - [email protected]

[email protected]
Create a post
You are not logged in. However you can subscribe from another Fediverse account, for example Lemmy or Mastodon. To do this, paste the following into the search field of your instance: [email protected]
  • 1 user / day
  • 1 user / week
  • 1 user / month
  • 1 user / 6 months
  • 0 subscribers
  • 55 Posts
  • 1 Comment
  • Modlog
  • mods:
  • Mazdak
  • UI: 0.18.4
  • BE: 0.18.2
  • Modlog
  • Instances
  • Docs
  • Code
  • join-lemmy.org