An operational framework

Stochastic tinkering occupies one end of this spectrum. Meanwhile, top-down design marks the other end. Given the relativity of space and time, as well as the realities of limited resources and non-negligible transaction costs (which cannot be ignored), we would be wise to investigate how to operate within this spectral framework.

Consider a serious disease, like AIDS or malaria, or, weigh the potential consequences of antibiotic resistance. These problems pose serious dilemmas to the future of our species, but how best to attack them is unclear: top-down design, bottom-up trial-and-error, or some combination of the two? Given Morton Meyers' splendid insights on this topic, via his book,Happy Accidents: Serendipity in Modern Medical Breakthroughs(Nassim Taleb says, "Read it twice."), it appears that top-down design has dominated research efforts for quite some time and that many of our marked advances in medical science (the discovery of penicillin, etc.) have emerged from someone engaged in top-down research who stumbled upon an unexpected development and then tinkered aggressively with this novel finding, progressing into previously uncharted territory. These "happy accidents" that emerge from the envelope of serendipity  represent the positive Black Swan hits that we desperately need in order to solve and mitigate pressing and very serious problems like our current AIDS epidemic.

So, then, should the prescription be to engage top-down design in research and attempt to "keep the door to the unknown ajar" (using Richard Feynman's message), waiting for serendipity to strike and then pursue these novelties to their logical conclusions? Or, should we tinker stochastically in search of solutions? Taking time into consideration, we must recognize that people are dying of AIDS (to continue this one example; there are many others) all the time, and time is of the essence. Does tinkering and trial-and-error solution searching fit into our time-constrained problem-solving efforts? Or, do we attack this problem from the top down, knowing that our backs are against the wall?

What should we do when time is of the essence and socially desirable results are needed (people's lives must be saved, emergencies, epidemics, etc.)?

The answer is far from clear, but I suggest we think in terms of hedged portfolios once again. Let's entertain the Barbell strategy one more time. In short, we could continue to engage top-down design efforts with 80% of our energy and resources, while the other 20% could be spread in a highly diversified manner across as many stochastic tinkering cheap health options as possible. This portfolio would gain exposure to the "happy accidents" that emerge as unexpected byproducts of top-down design, while also buying exposure to the field of stochastic tinkering options that, I suspect, elicits positive Black Swan hits at a much, much faster rate than top-down processes do. When we are pressed for time and serious social consequences are on the line, we would be wise to consider how to balance and blend bottom-up processes with top-down design. Bottom-up tinkering has the potential to save the day (especially when time is abundant and transaction costs are low), but our lives are short and we cannot delay acting to alleviate human suffering.

We need empiricism; we need to capture the ROI of this operational framework; and we need to experiment with different portfolios along this action spectrum.