Grading My 2018 Predictions For 2023

Scott grades his 2018 five-year predictions across AI, world affairs, US culture/politics, economics, science/tech, and x-risks (best call: AI generating images/stories to a prompt; worst: giving Roe's overturn only 1%), then lays out fresh 2028 predictions (the 'age of miracles and wonders' slow takeoff, action transformers, AI-made media, AI companions, Solow's Law). Honest calibration self-assessment plus a prescient-feeling AI section.

Quality 72 · Strong. Strong (upper): a valuable, honest calibration exercise that models good forecasting practice, with genuinely sharp AI foresight; kept mid-Strong because it is structurally a graded list of predictions plus new ones rather than a thesis-driven essay.

Claude’s paradigm shift 42 · Moderate. Moderate-low: prediction-grading is a recurring form for Scott and the content is topical; little novelty of idea.

Real-world impact 2 · Minor. Minor/within-discourse: models forecasting practice for the rationalist culture; no material real-world effect.