AI & ML

Handling long context still feels unreliable responses lose track

First post and replies  | Last post by Keith Bower, 6 years ago

User phrasing variation broke multiple prompt flows we designed

First post and replies  | Last post by Gina Pujals, 6 years ago

Prompt injection attacks recently came up in discussion how serious is it

First post and replies  | Last post by Kara Baker, 6 years ago

Minor bias observed in outputs not extreme but noticeable

First post and replies  | Last post by Tim Foss, 6 years ago

Outputs are technically right but not helpful in real scenarios

First post and replies  | Last post by Stian Ulriksen, 6 years ago

Feels like we are adding AI just to stay competitive not sure of real value

First post and replies  | Last post by Timothy Morrison, 6 years ago

Initial spike in usage then drop anyone else seeing this pattern

First post and replies  | Last post by Stella Khalif, 6 years ago

Model size decision still unclear smaller vs larger tradeoff is confusing

First post and replies  | Last post by Graeme Cochrane, 6 years ago

Accuracy is high on paper still users complain about answers

First post and replies  | Last post by David Manley, 6 years ago

Support automation didn’t reduce load as expected something feels off

First post and replies  | Last post by Diana Cunningham, 6 years ago