Silicon and Steel:: the Fusion Powering Next Industrial Age

Every automation wave before this one specialized. The ATM was extraordinary at dispensing cash and useless at anything else. The industrial robot arm was extraordinary at welding the same seam ten thousand times and useless the moment the part on the line changed shape. Specialization is what made these technologies predictable. You could look at a machine, understand exactly what it did, and reason confidently about which jobs sat downstream of it. What this book describes is not specialized in that way. The same underlying model that lets a machine read a paragraph and summarize it is now, with a different body attached, letting a machine look at a cluttered table and decide how to clear it. The cognitive and the physical are arriving from the same direction, on the same underlying technology, in the same handful of years. Economists do not have a settled model for that convergence, because it has not happened before. That absence of precedent is not proof of catastrophe. It is proof that the comforting history I described above may not apply cleanly, and that is reason enough to take the chapters that follow seriously rather than skeptically dismissing them.