Omar Shaikh is a Stanford PhD student, HCI and NLP researcher, and author of the award-winning UIST 2025 paper Creating General User Models from Computer Use. His work focuses on closing the human-AI grounding gap by helping systems develop a richer model of the people they assist.

In this episode, we explore better context for AI, the promise of General User Models (GUMs), calibration and confidence in user modeling, mixed-initiative interactions, and the design and privacy challenges of building systems that understand people more deeply.

Key Takeaways

• The Grounding Gap Matters: AI systems often fail not because they are weak, but because they lack the right context about the user.
• General User Models: GUMs aim to infer user goals, preferences, and state from computer-use behavior.
• Calibration and Confidence: Reliable user modeling requires not just predictions, but well-calibrated uncertainty estimates.
• Mixed Initiative and Privacy: More proactive AI systems can be powerful, but only if they respect privacy, ownership, and user control.

---

Timestamps

• 0:00 — Teaser
• 1:21 — Prelude: Introducing Omar Shaikh
• 2:07 — Monologue: Better Context for AI
• 4:22 — Bridging the Human-AI Grounding Gap
• 6:14 — Confidence scores in General User Models (GUMs)
• 7:32 — Calibration of General User Models
• 13:20 — Uses of General User Models
• 15:01 — Mixed Initiative Interactions
• 22:10 — Motivation for GUM
• 25:31 — Tabracadabra: tab everywhere!
• 27:01 — Design decisions in GUM
• 28:26 — Designing Interactive Experiences
• 32:11 — DITTO
• 33:06 — Work on domains without existing benchmarks
• 34:45 — Challenges of the GUM project
• 37:26 — Privacy and data ownership
• 38:57 — Finetuning a user model
• 44:09 — Mindblowing GUM inferences
• 49:02 — Social problems of GUMs
• 50:27 — GUM as a reflection tool

---

References

• Omar Shaikh’s research homepage
• Creating General User Models from Computer Use
• Tabracadabra
• Aligning Language Models with Demonstrated Feedback (DITTO)
• Principles of Mixed-Initiative User Interfaces
• Verification of Forecasts Expressed in Terms of Probability

---

Subscribe and follow us for more episodes!