MIRI's 2017 Strategy: Retaining Optionality on Alignment-Capabilities Insights
jessi_cata · x · 2026-08-16
jessicata cites MIRI's 2017 fundraiser and strategy update, noting that MIRI's strategy is closer to 'retaining optionality about which projects to help with our integrated alignment-capabilities insights on the task AGI / pivotal act path, which we pre-emptively keep secret' than 'we think <X org> should do a task AGI'. This responds to the debate about whether MIRI tried to take over the world, showing it retained optionality and secretly assisted.
More from AGI Musings
- Debate: Skepticism Over Anthropic's Claim of <2x AI R&D Acceleration — teortaxesTex · 2026-08-16
- AI could worsen institutional 'skin in the wrong game' purpose drift — xuanalogue · 2026-08-16
- Humanoids cheaper than cars could generate more value than employees — VraserX · 2026-08-16
- Anthropic accelerates biology and medicine efforts to cure cancer — teortaxesTex · 2026-08-16
- Anthropic CEO predicts AI could cure most human diseases in 5-10 years — Polymarket · 2026-08-16
- Expert Debate: Will Model Misalignment Cause the AI Bubble to Burst? — joshua_saxe · 2026-08-16