ForeCast [AI Narration] Design Sketches: Angels-on-the-Shoulder
Jan 28, 2026
Listeners hear sketches of five AI tools aimed at improving decision quality, including recommenders that optimize for long-term endorsement. They explore personalized learning tutors that adapt to goals and habits. The show outlines deep briefing systems that model context and stakeholders. It also covers reflection scaffolding for Socratic coaching and always-on guardian angels that nudge away from risky choices.
AI Snips
Chapters
Transcript
Episode notes
Decision Quality Gap Is Fixable
- People routinely make decisions that fall short of their considered values due to scarce attention and situational blind spots.
- Closing these gaps with supportive AI could substantially reduce costly personal and societal mistakes.
Optimize Feeds For Endorsed Value
- Build recommender systems that optimize for users' long-term endorsement rather than short-term engagement.
- Collect mixed metrics and occasional deep reviews to avoid naive optimization and learn what users value over time.
Design Adaptive Learning Tutors
- Create personalized learning tutors that map goals, assess gaps, and deliver spaced, adaptive curricula tied to users' routines.
- Start with lean variants or reusable resource maps to reduce upfront complexity and speed adoption.