Conversation

Notices

1. Embed this notice
   Cory Doctorow (pluralistic@mamot.fr)'s status on Tuesday, 02-Apr-2024 00:59:02 JST Cory Doctorow

   Here's a fun AI story: a security researcher noticed that large companies' AI-authored source-code repeatedly referenced a nonexistent library (an AI "hallucination"), so he created a (defanged) malicious library with that name and uploaded it, and thousands of developers automatically downloaded and incorporated it as they compiled the code:

   https://www.theregister.com/2024/03/28/ai_bots_hallucinate_software_packages/

   1/

   • gidi likes this.
   • Embed this notice
     Haelwenn /элвэн/ :triskell: (lanodan@queer.hacktivis.me)'s status on Tuesday, 02-Apr-2024 07:40:14 JST Haelwenn /элвэн/ :triskell:

     in reply to
     @pluralistic
     > Lanyado also said that there was a Hugging Face-owned project that incorporated the fake huggingface-cli, but that was removed after he alerted the biz.
     
     Wonderful…
   • Embed this notice
     Pēteris Krišjānis (peteriskrisjanis@toot.lv)'s status on Tuesday, 02-Apr-2024 07:41:09 JST Pēteris Krišjānis

     in reply to

     • Christoffer Vig

     @Babadofar @pluralistic no, we cannot, what are you even about. It is just fricking random choice sorting hat.

     Haelwenn /элвэн/ :triskell: likes this.
   • Embed this notice
     Christoffer Vig (babadofar@snabelen.no)'s status on Tuesday, 02-Apr-2024 07:41:10 JST Christoffer Vig

     in reply to

     @pluralistic thanks for sharing. That's a very good summary explanation of the challenges of AI /Llm.
     It's going back to the question of what is truth. Can we trust AI with important decisions.