Research-backed, actionable articles on AI at work, endurance training and how people learn. Every claim has a source.
Rinsing maltodextrin and spitting it out cut a 1-hour cycling trial by 1.8 minutes. Pooled across 35 studies the gain falls to 0.18 SD. The dose and the timing.
Read the article →Models scored 39% lower when a task arrived turn by turn instead of in one brief. Four studies on multi-turn drift, and the restart rule I teach.
Read the article →Surgical teams missed 6% of critical steps with a checklist and 23% from memory. The trials, one failed 101-hospital rollout, and how to write one that gets finished.
Read the article →Prompted self-explanation averages g = 0.55 across 64 reports and g = 0.46 in digital settings, and zero on plain facts. The distinction, and a protocol.
Read the article →Two meta-analyses and three trials on HRV-guided endurance training: the same fitness off fewer hard sessions. Measurement rules and the daily decision rule.
Read the article →The best agent completed 30.3% of 175 simulated office tasks, and its consistency across 8 runs fell to 25%. Four benchmarks, and the protocol I teach.
Read the article →Nine practice problems scored no better than three after 4 weeks, while the same reps split across two sessions doubled retention. A stopping rule with numbers.
Read the article →A 10% faster step rate cuts peak patellofemoral force by 14% in the lab. The nine studies that bound the effect, the cost in speed, and a 6-week plan.
Read the article →Word error rates near 15%, sentences nobody spoke, summaries wrong 30.4% of the time. What 6 studies measured, and the review to run before forwarding.
Read the article →A PNAS trial with nearly 1,000 students: GPT-4 lifted practice scores 48%, then cut exam grades 17% once access went away. Four studies and a protocol.
Read the article →