My view is that forecasting in general is a weak concept. Our science and technology do not deal well with complex systems, and in some important ways they never can; we’ll never be able to measure initial conditions accurately enough, we’ll never be able to simulate accurately enough, we cannot escape chaos, and the more compute we throw at it the more complex the whole intertwined system becomes—because the compute is part of the system.
But we have an exemplar of what we can do instead; biological evolution. It is pure empirical science, no theory or prediction required. It is our existence proof that a complex context can be navigated successfully… depending on your definition of success. I suggest that we should crank up our humility to more realistic levels and adopt a more stochastic approach; ie try lots of randomized, low-risk experiments in parallel and leave the analysis until set the solution emerges.
The most important avenue of scientific inquiry is that of understanding the necessary conditions for a functional memetic evolutionary process. Given how far we have drifted from the original basis of human culture (many small groups, with low mutual bandwidth and strong life/death feedback) it seems quite plausible that our current system (a globally integrated and highly centralized civilization with increasingly abstract and recursive feedback mechanisms) is selecting strongly for worse ideas as time progresses.
My view is that forecasting in general is a weak concept. Our science and technology do not deal well with complex systems, and in some important ways they never can; we’ll never be able to measure initial conditions accurately enough, we’ll never be able to simulate accurately enough, we cannot escape chaos, and the more compute we throw at it the more complex the whole intertwined system becomes—because the compute is part of the system.
But we have an exemplar of what we can do instead; biological evolution. It is pure empirical science, no theory or prediction required. It is our existence proof that a complex context can be navigated successfully… depending on your definition of success. I suggest that we should crank up our humility to more realistic levels and adopt a more stochastic approach; ie try lots of randomized, low-risk experiments in parallel and leave the analysis until set the solution emerges.
The most important avenue of scientific inquiry is that of understanding the necessary conditions for a functional memetic evolutionary process. Given how far we have drifted from the original basis of human culture (many small groups, with low mutual bandwidth and strong life/death feedback) it seems quite plausible that our current system (a globally integrated and highly centralized civilization with increasingly abstract and recursive feedback mechanisms) is selecting strongly for worse ideas as time progresses.