Comment

Comment on To explore AI bias, researchers pose a question: How do you imagine a tree?

vacuumflower@lemmy.sdf.org ⁨2⁩ ⁨days⁩ ago

Thus, a user receives an answer that has already undergone a filtering of sorts.

Wouldn’t this be a predictable trait of a system predicting next most likely token based on lossy compression of specific datasets and other lossy optimization?

source

Sort:hotnew top

Eq0@literature.cafe ⁨2⁩ ⁨days⁩ ago
Depends. For an expert, that is self evident (even if it might not be clear which biases have been incorporated). But that is not how it has been marketed. Chatgpt and similar are perceived as answering “the truth” at all times, and that skews the user’s understanding of the answers. Researching how deeply the answers are affected by the coders’ bias is the focus of their research and a worthwhile undertaking to avoid overlooking something important

source
- spankmonkey@lemmy.world ⁨1⁩ ⁨day⁩ ago
  
  For an expert, that is self evident
  
  I am far from an expert, but it seemed obvious to ne.
  
  source
  - Eq0@literature.cafe ⁨1⁩ ⁨day⁩ ago
    I teach, nothing is evident to anyone 😭
    
    source