The Learning+ Podcast Ep. 36: Learning, Systems, and the Design of Work with Michael Snelgrove (Part 1)
10 snips
Feb 15, 2026 
Michael Snelgrove, a behavioural scientist with frontline experience in high-stakes settings, discusses how learning and performance function under pressure. He explains using choice architecture to improve decisions, why training can backfire, the limits of self-directed learning, and how system design, mental models, and better feedback help people perform when it matters most.
AI Snips
Chapters
Books
Transcript
Episode notes
Reducing Options Saved Lives In Afghanistan
- Michael Snelgrove recounts using behavioral design in southern Afghanistan to improve roadside bomb detection without relying on widespread training.
- He read Nudge at night, applied choice-architecture fixes, and iteratively reduced options to simplify requests for intelligence capabilities.
Choice Overload Beats Expertise In High Stakes
- Too many choices without domain knowledge degrades decisions, like giving customers an ingredient list instead of curated menu options.
- In Afghanistan they replaced long option forms with simple if-then rules and estimated ~30% detection improvement.
Well Intentioned Programs Can Do Harm
- Solutions in HR and L&D can cause harm (iatrogenics) when poorly designed or deployed at scale.
- Michael links this risk to unconscious bias training rollouts that sometimes increase bias despite good intentions.
