The Practical and Theoretical Future of Computation in Physics

I started using computers for physics around 1972, and for the past 30+ years I've been building things like Mathematica and Wolfram|Alpha that have been used to do a lot of physics. Around 1980 realized that computation can be used not only for the practical doing of physics, but also as a framework for thinking about physics. Among other things that led to my book A New Kind of Science, and by now many new models in many fields (including physics) are getting made in terms of programs rather than traditional mathematical constructs.

In this talk, I'll discuss the framework I've been building for doing physics by computer, and its future implications—both in terms of theoretical modeling, data analysis and machine learning, as well as connections to the knowledgebase we've accumulated, and to real-time data taking. I'll also discuss what it means to think about the foundations of physics in computational terms—the implications of phenomena like undecidability and computational irreducibility, and what one can anticipate about the theoretical structure of physics in the future.