Users criticized the post on AI endpoints shifting from direct model access to control systems as intentionally misleading and alleged Trump administration interference in AI classifiers.
Based on 2 visible X reactions from 3 accounts; directional sample.
Ask a question below.
Published answers will appear here.
@elder_plinius Your OP was intentionally misleading. I don’t like it
@elder_plinius Many people actually believe the model is volitionally "refusing" requests when the classifier activates, in a large part due to Anthropic's misleading notice messages. I think it's a nasty thing to falsely insinuate & you shouldnt feed the misconception. https://x.com/repligate/status/2073824932635799914?s=20
@repligate not intentionally, but could have phrased it better. was operating under the context that the system we refer to as “Fable” is really Mythos in a straitjacket surrounded by overly strict wardens manipulating comms
@repligate of course, and it’s a real shame that most endpoints don’t connect us to models anymore but systems of control
Users criticized the post on AI endpoints shifting from direct model access to control systems as intentionally misleading and alleged Trump administration interference in AI classifiers.
Based on 2 visible X reactions from 3 accounts; directional sample.
Ask a question below.
Published answers will appear here.