Conversation

Notices

1. Embed this notice
   Prof. Emily M. Bender(she/her) (emilymbender@dair-community.social)'s status on Monday, 17-Feb-2025 06:41:36 JST Prof. Emily M. Bender(she/her)

   I'm frequently struck by the phrase "unintentional bias" which comes up frequently in work on LLMs. Whose intentions are in question? Why do we have to absolve those people all the time?

   • Rich Felker repeated this.
   • Embed this notice
     Nicole Parsons (npars01@mstdn.social)'s status on Monday, 17-Feb-2025 12:02:32 JST Nicole Parsons

     in reply to

     @emilymbender

     The developers of LLM's & AI have been warned about bias for over a decade yet the industry makes strong efforts to purge the harbingers.
     https://www.rollingstone.com/culture/culture-features/women-warnings-ai-danger-risk-before-chatgpt-1234804367/

     https://www.wired.com/story/google-timnit-gebru-ai-what-really-happened/

     Why?
     Purges of dissenting staff occur because the investors behind these initiatives want the results so badly, they're prepared to fund it limitlessly, no matter its poor quality.
     https://www.theregister.com/2025/02/12/larry_ellison_wants_all_data/

     https://www.forbes.com/councils/forbestechcouncil/2024/02/02/artificial-intelligence-the-new-eyes-of-surveillance/

     https://www.theregister.com/2024/09/16/oracle_ai_mass_surveillance_cloud/

     Silicon Valley is more interested...
     1/2