35
Testing suggests Google's AI Overviews tell millions of lies per hour (arstechnica.com)
submitted 7 months ago by to c/world
2 comments fedilink hide all child comments
 

**

90% of the time right means 10% of the time wrong, huge deal when you deal with billions of queries! **

you are viewing a single comment's thread
view the rest of the comments
[–] 3 points 7 months ago* (last edited 7 months ago)

This isn't limited to Google's AI but true for all of LLMs. But as we all know, they are not 'lying' but rather just repeating the 'lies' they have learned.

  • permalink
  • fedilink
  • source
    •