Regina Dugan, 2012

“What would you attempt to do if you knew you could not fail?”

Then-DARPA director Regina Dugan used this question in her 2012 TED talk about ambitious research programs. She presented projects such as a robotic hummingbird and advanced prosthetics to show what disciplined teams can attempt when fear of failure does not define the boundary. The question expands imagination, but Dugan’s examples also depended on technical rigor, milestones, and a portfolio of experiments. Ambition becomes productive when a bold outcome is paired with learning that can survive individual setbacks. The enduring management question is which decision the idea changes and what evidence would show that the interpretation is wrong. Managers should describe the causal chain they expect: the action, the immediate response, and the outcome worth improving.

Businesses can frame a consequential goal, fund several approaches, and review them against evidence rather than prestige. Teams should be able to stop a path early, share what failed, and redirect resources without treating every setback as a career event. A useful principle should improve a choice before it becomes a retrospective slogan. A review date protects the idea from drifting into a permanent initiative that nobody can evaluate or end.

Some failures are unacceptable when they threaten safety, rights, or public trust. Confidence must not erase uncertainty or oversight. High-risk experimentation needs bounded environments, independent review, and clarity about who bears the consequences. That requires attention to incentives, affected people, and the conditions under which the idea stops applying. It should include the strongest credible objection, not only evidence most favorable to the proposal.

Organizations often speak about innovation while rewarding only predictable delivery. Dugan’s question exposes that contradiction and invites leaders to build conditions where responsible attempts at difficult work are possible. Ask a team to name a goal it avoids because failure would be punished. Define a safe experiment, evidence threshold, stop rule, and learning review, then protect the team when it reports an unwelcome result promptly. Recording the decision and its limits gives future teams knowledge they can inspect rather than a success story they are asked to repeat. That habit strengthens trust because people can see that learning, not defending the original author, is the purpose of measurement. The strongest evidence will combine a measurable outcome with a clear account of how the result was produced. Numbers show scale and direction, while interviews, observations, and exception logs reveal why the pattern occurred. Using both protects the organization from chasing a convenient average and helps it identify which customers, roles, or locations need a different design before the practice expands. The owner should return with evidence and a recommendation at the agreed review date.