[D66] AI: GPT-5 Risks

[René Oudeweg]

https://www.youtube.com/watch?v=JyVH4FbSwFo

GPT-5 Presents EXTREME RISK (Google's New Warning)
TheAIGRID
75,4K abonnees
363.220 weergaven 10 jun 2023 #Robotics #AI #chatgpt


--

Ref:

Orca: Progressive Learning from Complex
Explanation Traces of GPT-4
https://arxiv.org/pdf/2306.02707.pdf

Predictability and Surprise in Large Generative Models
https://arxiv.org/pdf/2202.07785.pdf

Model evaluation for extreme risks
https://arxiv.org/pdf/2305.15324.pdf#page=17&zoom=100,46,109

AI suggested 40,000 new possible chemical weapons in just six hours
https://www.theverge.com/2022/3/17/22983197/ai-new-possible-chemical-weapons-generative-models-vx
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.tuxtown.net/pipermail/d66/attachments/20230620/4a1a1c83/attachment-0001.html>