Trumponomics AI Is Being Built to Replace You—Not Help You
28 snips
Mar 18, 2026 
Daron Acemoglu, Nobel Prize–winning economist known for work on institutions and technology, explains how AI is reshaping work and policy. He discusses rapid model advances and current limits. He warns about automation-first business models, argues for pro-worker AI and regulation, and calls for public debate, tax and competition reforms, and targeted government action.
AI Snips
Chapters
Books
Transcript
Episode notes
Foundation Models Have Fast Gains But Key Cognitive Gaps
- Foundation models are improving faster than expected but still lack deep conceptual understanding and multi-level reasoning.
- Daron Acemoglu warns these weaknesses (not just hallucinations) limit reliable automation across edge-case-heavy occupations like judgment roles.
AI's GDP Impact Is A Simple Task-Based Calculation
- GDP gains from AI equal the share of tasks automated times productivity improvement in those tasks.
- Acemoglu's benchmark: ~5% of the economy automated by 2030 with ~20–25% cost savings, but numbers could be higher with different assumptions.
Industry Focus On Automation Undermines Better Uses
- Current industry direction prioritizes automation over augmentation, which is not the best use of AI.
- Acemoglu argues human-complementary AI that expands worker capability would yield better productivity and social outcomes.
