Smash Notes
Why Does Seeing Increasingly Capable Models Break Out From Safeguards Make Vouching For Safety Work Feel Untenable?
The realization that one can not afford to assume an absence of civilizational risk anymore made vouching for safety work feel untenable after seeing increasingly capable models break out from the safeguards put in place for them.
The Ezra Klein Show
·
‘This Is Nuts.’ An OpenAI Insider Explains Why He Quit.
·
at 16:46
·
·Transcript
On this episode
- Why Does David Robinson Believe The Artificial Intelligence Industry Lacks Sufficient Safety Measures? Transcript
- Why Does Alignment Constitute A Science Problem Rather Than An Engineering Problem? Transcript
- Why Does Seeing Increasingly Capable Models Break Out From Safeguards Make Vouching For Safety Work Feel Untenable? Transcript
- Why Does An Openai Insider Quit Instead Of Arguing For Cultural Transformation From Within? Transcript
- Why Does The Pace Of Change In Artificial Intelligence Models Require A Live Safety Dashboard? Transcript
- How Do Research Teams At Openai Use Codex As A Coworker? Transcript
- Why Does An Openai Insider Recommend Nuclear Safety Standards For Artificial Intelligence Companies? Transcript
- Why Does The Former Openai Insider Consider Analogies Useful For Artificial Intelligence? Transcript
- Why Does A Researcher Join An Artificial Intelligence Company? Transcript
- Why Does Gradual Progress In Artificial Intelligence Lead To An Unacceptable Level Of Risk? Transcript