Institutional & Societal Implications¶

The broader impact of AI on academia, research institutions, and society.

Summaries¶

AI as Normal Technology — Narayanan & Kapoor: AI as slow-diffusing technology, not revolution
The Bitter Lesson — Sutton: general computation-scaling methods always win
The Shape of AI — Mollick: jagged frontier, bottlenecks, and salients
Why No One Can Agree About AI Progress — Kim's Mind/Body/Instructions model explaining why expert opinion splits

Academics Need to Wake Up on AI — Kustov on urgency of AI adoption in academia
Academics Need to Wake Up, Part II — Kustov's follow-up with concrete recommendations
Academics Need to Wake Up, Part III — Kustov: human slop predates AI; not using AI is malpractice
AI is Already 10x-ing Academic Research — Hall (Stanford): vote-by-mail auto-replication, 100x institution proposal
AI, Ideas, and the Future of Research — Kansoy (Oxford): six-section deck arguing judgment is the new scarce input; LLM + SKILL.md = Agent; full wealth-composition demo
Some Thoughts on AI and Research — Andrews (MIT): case-based framework for PhD skill investment under AI uncertainty
The Train Has Left the Station — Messing & Tucker (Brookings) on agentic AI and social science
Chris Blattman X Post on Claude Code Adoption — Blattman's social media post on adopting Claude Code

Zero Profit Condition Is Coming — Cunningham on competitive equilibrium, gains for least experienced
Attention, Verification and Congestion — Productivity externalities, convex costs of faster work
Faculty Adoption of AI — AI agents as experience goods, security risks, subsidies needed
Hadn't the Satisfaction Always Been in the Discovering — Intrinsic joy of research vs AI automation

Something Big Is Happening — Shumer on imminent AI disruption
What AI Got Wrong — Blattman on plausible-but-wrong AI errors requiring domain expertise
The AI People Have Been Right a Lot — Matthews's public 2015 update; track record of AGI-pilled predictors
Agentic AI and the Next Intelligence Explosion — Evans/Bratton/Agüera y Arcas: plural intelligence; society of thought inside reasoning models; institutional alignment vs. RLHF