In this talk, I explore how scientific breakthroughs over the last four centuries have reshaped not only our understanding of the natural world, but also how we think, organize, and act within it.
I begin with the distinction between clock problems and cloud problems, a framing inspired by Karl Popper. Clock problems can be extraordinarily complicated, yet they are ultimately solvable. Cloud problems, by contrast, are adaptive, shaped by human behavior, and resistant to permanent solutions. Much of the modern world, I argue, was built using clock-based thinking, an approach profoundly influenced by Newtonian mechanics and its emphasis on determinism, prediction, and control.
I then trace how successive scientific revolutions, from thermodynamics and evolution to quantum mechanics, computation, and artificial intelligence, challenged the assumptions underlying that worldview. These revolutions revealed a reality that is probabilistic rather than deterministic, emergent rather than reducible, and deeply interconnected. Many of today’s most pressing challenges arise when we apply clock thinking to cloud problems.
In the final part of the talk, I turn to AI and creativity. AI is extraordinarily powerful as a tool for prediction and combinatorial creativity, but it remains bounded by the assumptions embedded in its training data. Genuine breakthroughs, I argue, require something fundamentally human: the ability to question assumptions, reframe problems, and remain productively puzzled by what others take for granted. Moving from clocks to clouds is not about abandoning rigor, it is about expanding our ways of thinking to better navigate a complex world.
We now live in a world that demands comfort with uncertainty, adaptability, and plural perspectives. We must stop thinking only like clockmakers—perfecting mechanisms—and learn to think like cloud-watchers: seeing patterns as they form, fostering direction without rigidity, and finding possibility in uncertainty.
