Thiel on the pace of innovation and planning – gradualism, punctuated equilibria and Small Plans

During the recent, rainy weekend we spent in Seattle I did some reading. One of the things that I came across was a fascinating interview by Fukuyama with Peter Thiel. A couple of different things in the interview struck me as worth reflecting on. The first and most well-known is Thiel’s view that the pace of innovation has slowed down and that we are seeing massive deceleration of technological change.


So you have these two different blind spots on the Left and Right, but I’ve been more interested in their common blind spot, which we’re less likely to discuss as a society: technological deceleration and the question of whether we’re still living in a technologically advancing society at all. I believe that the late 1960s was not only a time when government stopped working well and various aspects of our social contract began to fray, but also when scientific and technological progress began to advance much more slowly. Of course, the computer age, with the internet and web 2.0 developments of the past 15 years, is an exception. Perhaps so is finance, which has seen a lot of innovation over the same period (too much innovation, some would argue).

There has been a tremendous slowdown everywhere else, however.

There seems to be a number of different questions here that are slightly different. One is if the pace of technological change is slowing down, another if the pace of innovation is slowing down. Technological change is not necessarily equal to innovation, depending on your definition of innovation. If we use the Schumpeterian definition of innovation as ideas brought to a market, then the pace of technological change is not enough: we need to look at new ways of making raw materials available, new ways of organizing ourselves, new products, new services and a lot more. In order to measure all of this our measurements also need to be able to take into account new values produced, something that is of course notoriously difficult.

So how could we determine the answer to any of these questions? It looks as if the answer is the same for most versions of the problem. We decide on a basket of numbers and track that basket from an arbitrary point in history and plot it. If innovation (or technical change) is accelerating we expect to see a curve that proves that, or we expect to see a flatline if we believe with Thiel that it is actually decelerating. The construction of those baskets will matter a lot. On transportation for example we could measure safety: number of people killed per mile travelled, or we could measure energy consumption per mile travelled or we could measure lots of other things. There will always be a set of data sets such that it implies either deceleration or acceleration. The debate will then be about what basket looks most reasonable, and that will most likely be a value judgement in the end.

But there is a third possibility. It could well be that the pace of change is not continuous, but that what we are seeing in innovation is much more like the punctuated equilibria that evolutionary biologists argue we see in evolution. Not incremental, continuous change, but stepwise change. Punctuated equilibria would look very much like deceleration, but the predictions for the two scenarios are very different. And the solution sets look very different too.

If we do believe in continuous change (or gradualism) in innovation we will look for things to speed change up again, boosts so that we can accelerate again. If we believe in punctuated equilibria we will look for things that will set of cascades of innovation, disruptive innovations that will shift the equilibrium. The policy prescriptions will also be very different: if we believe in continuous change we will invest more in existing institutions and encourage more R&D. If we believe in punctuated equilibria we will try to ensure that incumbents are not protected and that we invest in moon shots. Roughly.

So from the the same data sets that imply deceleration we could infer punctuated equilibria. I think the deceleration meme is destructive, and believe in the equilibria meme, but I guess my big problem is to devise an experiment that could be used to determine which scenario we are in from the data that we are gathering. That is an interesting problem.


Another part of the interview that I found eye-opening and that rang very true with me was Thiel’s observations on planning:

 We have different kinds of challenges on the government side. One is a little more philosophical in nature: We tend to think the future is indeterminate. But it used to be seen as a much more determinate thing and subject to rational planning. If it’s fundamentally unknowable, it doesn’t make sense to say anything about it. To put it in mathematical terms, we’ve had a shift from thinking of the world in terms of calculus to statistics. So, where we once tracked the motions of the heavenly bodies and could send Voyager to Jupiter over a multiyear trajectory, now we tend to think nature is fundamentally driven by the random movements of atoms or the Black-Scholes mathematical model of financial markets—the random walk down Wall Street. You can’t know where things are going; you only know they’re going to be random. I think some things are true about this statistical view of the future, but it’s extremely toxic for any kind of rational planning. It’s probably linked in part to the failure of state communist central planning, though I would argue that there is something to be said for some planning over no planning. We should debate whether it should be decentralized or centralized, but what the United States has today is an extremely big government, a quasi-socialist government, but without a five-year plan, with no plan whatsoever.

I think this is right. Complexity, information deluge and the general failure of Big Planning has made us ditch even what we could call Small Planning. But that is lethal. Without any plan you are doomed to just be the victim of circumstance, adrift at the sea of possibility. This means that Government creates very little in terms of selection pressure on innovation or the economy. We should not confuse the construction of selection pressure with planning, by the way. If we set up prizes for certain innovations, if we set up the system to reward certain kinds of problem solving we are not planning for how those problems will be solved. We are not engaging in the Big Planning of the Soviet Union. The lack of any kind of planning essentially guarantees that the evolution of our society now is in a state of simple evolutionary drift. And drift does not create adaptations to the complexity we are facing.

Small Planning is setting goals that are audacious but measurable, having an idea of what the story is (SMART goals). Thinking about the change we want in micro terms. Not managing the whole complex system of society, but organizing to change the bits and pieces we know. Small Planning is the process of extremely detailed thinking about these changes, and continuous evaluation.

Whenever you have a big problem you need to start with a Small Plan…

Be Sociable, Share!

Kommentera gärna!