Dan Saffer
Dan Saffer is an Assistant Professor of the Practice at the Human-Computer Interaction Institute at Carnegie Mellon University, a product design leader, and the author of four books: Designing Devices (2011), Designing Gestural Interfaces (2008), Designing for Interaction (2006, 2009) and the best-selling Microinteractions (2013).
Since 1995, he’s designed devices, apps, websites, wearables, appliances, automotive interiors, services, social networks, and robots. He’s worked at and for such companies as Twitter, Smart Design, Samsung, Jawbone, CNN, Philips, and Microsoft — and was most recently Head of Product Design at Flipboard.
He graduated from Carnegie Mellon with a Master’s in Design, Interaction Design.
Dan's Featured Talks
The New Era of AI UX
For decades, UX designers have built “hammers”—efficient, predictable tools designed for users to click, drag, and accomplish specific tasks. The rise of AI requires a fundamental shift in mindset: we are no longer designing static tools, but dynamic “apprentices” that can collaborate, suggest, and act on our behalf. This keynote explores the transition from Traditional UX to AI UX, defining the new roles users must play and how to design interfaces that support them.
From this session, attendees will:
- Shift your mindset: Move from designing for users “doing things” to designing for users “directing and monitoring”.
- Identify the 4 AI User Modes: Learn to design specifically for the four distinct behaviors of AI interaction: Directing, Monitoring, Intervening, and Approving.
- Embrace Ambiguity: Understand why AI interactions are uniquely difficult due to ambiguity in both inputs and outputs, and how to design around it
Dan will unpack:
- The Hammer vs. The Apprentice: A Metaphor for Understanding the Difference Between Tools with No Personality or Bias vs. Proactive, Collaborative Partners.
- Traditional UX vs. AI UX: Contrasting linear workflows (“click here, go there”) with conversational, fuzzy, and open-ended interactions.
- The “Make the Right Thing / Make the Thing Right” Model: A dual approach to validation, ensuring you are solving the right user intent before refining the interaction mechanics.
- Defining Delight in AI: Moving beyond visual polish to finding delight in anticipation, insight, empowerment, and graceful collaboration.
The Art of Human-AI Collaboration
The dream of AI “full autonomy” is often a UX nightmare, leading to errors, lack of trust, and skill degradation. The sweet spot of AI design lies in “Human-AI Collaboration”—keeping the human in the loop at the right moments. This technical and strategic keynote provides a taxonomy of design patterns for breaking down complex AI tasks into manageable steps where humans can intervene, correct, and approve.
From this session, attendees will gain:
- Avoid the “Sparkle Button” Trap: Learn why “all-or-nothing” AI interactions fail and how to break tasks into discrete moments for human oversight.
- Design for Intervention: Implement specific UI patterns like “Return of Control” and “Collaborative Planning” to give users agency over the AI.
- Build Trust Through Friction: Discover when to intentionally slow down the process (Cognitive Forcing Functions) to prevent overreliance and anchoring bias.
Dan will unpack:
- Taxonomy of Intervention: A set of 7 mechanisms for human control, including User Confirmation, Granular Control, and Explicit Dialogue.
- Adaptive UI (AUI): A process for automating repetitive tasks based on user habits and confidence levels.
- Cognitive Forcing Functions: Design choices that prompt users to think critically rather than blindly trust the AI, such as withholding assistance when uncertain.
- Trust & Explainability Model: Understanding that trust is based on mental models, and using “WAIST” (Why Am I Seeing This) to make AI behaviors transparent.
