OpenAI whistleblower interview sparks debate: AI ethics vs alignment framing
examachine · x · 2026-09-12
Commenting on the BBC interview with OpenAI whistleblower Daniel, the author argues that alignment is a flawed, behavioral framing that ignores the root problem: some humans are demonstrably evil. He contends that alignment therefore produces evil outcomes and calls instead for unbreakable "EI Ethics" — an unbreakable ethical framework for AI rather than behavioral alignment. A pointed, if thinly argued, critique of the alignment paradigm.
More from AGI Musings
- Sam Altman reportedly tells OpenAI staff that slowing AI development is on the table — VraserX · 2026-09-12
- Prompt Engineering Is Over — Context Engineering Is the Real Meta — JafarNajafov · 2026-09-12
- World Models: The Next AI Frontier Is Reasoning About the Physical World — ingliguori · 2026-09-12
- AI Futures Researcher Admits 'Aligned' AI Really Serves a President or CEO — Then Flinches — victor_explore · 2026-09-12
- First Full Month of AI Overviews in France: All Top 10 News Sites Lost 11-30% of Search Traffic — gaganghotra_ · 2026-09-12
- Commenter Argues Musk's 10-20% Doom Odds Actually Imply 100% Conditional on ASI — JOBhakdi · 2026-09-12