Dan Saffer
Dan Saffer is 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, Microsoft and Flipboard. He is currently Assistant Professor of The Practice and Associate Director of Outreach at the Human-Computer Interaction Institute at Carnegie Mellon.
Presentations
Leadership By Design 2026
Six Futures
In an AI world, apps don’t have a single future. They have six, and soon, a given product might occupy three or four of those positions at the same time. For companies building AI-powered products, understanding this spectrum is a strategic imperative.
We are at an inflection point where the traditional app — the thing you open, navigate, and close — is becoming just one of several ways users interact with what a product does, and in the future, possibly the least common one. Each of these futures carries a different business model, a different relationship with the user, and a different answer to the question of what makes a product defensible.
Two of these futures are already taking shape in prototyping work from my UI for AI research team at Carnegie Mellon. The first explores what happens when an operating system understands what you’re working on, not just which apps are open: assembling cross-application workspaces around tasks, generating parallel approaches so users compare real outputs before committing, and preserving the exploratory thinking that current AI tools collapse into a single generated answer. The second explores ephemeral interfaces — AI that surfaces within whatever tool you’re already using at the moment it’s relevant, then dissolves when the work is done, earning every appearance rather than demanding permanent screen real estate. Both represent a fundamental shift from interfaces organized around applications to interfaces organized around what someone is actually trying to accomplish. This talk will show what that shift looks like in practice.
The spectrum also exposes a deeper shift in how products get built. AI is collapsing the traditional roles on product teams. When a PM can generate mockups, a designer can write working code, and an engineer can produce competent UI, the artifact is getting commoditized. What remains durable is judgment: knowing which of twenty AI-generated options actually works in context, knowing what good looks like in a domain where the consequences of getting it wrong are regulatory and reputational, and knowing which friction to preserve rather than eliminate. That last question turns out to be the central design challenge of the moment. When you remove all friction from an activity, you don’t just make it efficient — you hollow it out. Automating the struggle away doesn’t create value. It destroys it.
This talk draws on a year of research, prototyping, and teaching at Carnegie Mellon’s Human-Computer Interaction Institute to map where AI products are headed and what it means for the people building them.
