Safety Researcher Critiques Superintelligence Strategy: Jobs and Control Arguments Flawed
DKokotajlo · x · 2026-08-11
AI safety researcher Ryan Greenblatt sharply criticized current strategies for handling superintelligence.
He argued that while aligning AIs with individuals and preventing corporate monopolies on advanced AI is positive, the current vision for managing superintelligence fails to address the biggest problems:
- Jobs: Arguments that labor will retain a significant fraction of value long-term (such as 'finite compute' or 'AIs will create higher value work') do not make sense.
- Control: The assumption that humans may retain most control over the future is also challenged.
He emphasized that the current strategy seems unserious about resolving these issues.
More from AGI Musings
- AI Safety: Why Compute Limits Won't Work as a Stopping Point for AI Development — scottleibrand · 2026-08-11
- Chollet's Deep Dive: Future AI Architectures Will Pivot to Symbolic Learning — burny_tech · 2026-08-11
- Rethinking LLM Usage: Don't Assume Every Problem is Solvable — zeeg · 2026-08-11
- Compute Constraints and Closed Frontiers: MIT Tech Review on the Dilemma of Academic AI Research — nordicinst · 2026-08-11
- Doctor to AI firms: give us unlimited tokens, we'll give you a cancer-free world — DrDatta_AIIMS · 2026-08-11
- Reasoning Models Plus Multi-Agent Swarms Will Cause Exponential Token Consumption — robleclerc · 2026-08-11