Rock star data scientist Nate Silver wrote a long article on meteorology for the New York Times:
Why are weather forecasters succeeding when other predictors fail? It’s because long ago they came to accept the imperfections in their knowledge. That helped them understand that even the most sophisticated computers, combing through seemingly limitless data, are painfully ill equipped to predict something as dynamic as weather all by themselves. So as fields like economics began relying more on Big Data, meteorologists recognized that data on its own isn’t enough.
Technology Review covers Stuart Kauffman‘s work to find a mathematical model for autocatalytic sets, the process by which life may emerge from molecules:
What makes the approach so powerful is that the mathematics does not depend on the nature of chemistry–it is substrate independent. So the building blocks in an autocatalytic set need not be molecules at all but any units that can manipulate other units in the required way.
These units can be complex entities in themselves. “Perhaps it is not too far-fetched to think, for example, of the collection of bacterial species in your gut (several hundreds of them) as one big autocatalytic set,” say Kauffman and co.
And they go even further. They point out that the economy is essentially the process of transforming raw materials into products such as hammers and spades that themselves facilitate further transformation of raw materials and so on. “Perhaps we can also view the economy as an (emergent) autocatalytic set, exhibiting some sort of functional closure,” they speculate.
Could it be that the same idea–the general theory of autocatalytic sets–can help explain the origin of life, the nature of emergence and provide a mathematical foundation for organisation in economics?
I find this very interesting, but don’t get too excited. These sorts of grand unification theories are extremely elusive. I’m also skeptical of these sorts of models which try to find universal rules for all types of systems.
Great new interview with Nassim Taleb by one of his former teachers at Wharton:
Taleb: The events in the Middle East are not black swans. They were predictable to those who know the region well. At most, they were gray swans or perhaps white swans. One of the lessons of “Wild vs. Mild Randomness,” my chapter with Benoit Mandelbrot in your book, is what happens before you go into a period of wild randomness. You will find a long quiet period that is punctuated with absolute total turmoil…. In The Black Swan, I discussed Saudi Arabia as a prime case of the calm before the storm and the Great Moderation [the perceived end of economic volatility due to the creation of 20th century banking laws] in the same breath. I was comparing Italy with Saudi Arabia. Italy is an example of mild randomness in comparison with Saudi Arabia and Syria, which are examples of wild randomness. Italy has had 60 changes in regime in the post-war era, but they are inconsequential…. It is a prime example of noise. It’s very Italian and so it’s elegant noise, but it’s noise nonetheless. In contrast, Saudi Arabia and Syria have had the same regime in place for 40 some years. You may think it is stability, but it’s not. Once you remove the lid, the thing explodes.
The same kind of thing happens in finance. Take the portfolio of banks. The environment seemed very placid — the Great Moderation — and then the thing explodes.
Herring: I would agree that people knew the Middle East was very vulnerable to turmoil because of the demographics, a very young population, and widespread unemployment, the dissatisfaction with the distribution of income and with regimes that were getting geriatric. But knowing how it would unfold and knowing that somebody immolating themselves in a market in Tunisia would lead to this widespread discontent — and we still don’t know how it will end — is a really remarkable occurrence that I think would be very difficult to predict in any way.
Taleb: Definitely, and it actually taught us to try not to predict the catalyst, which is the most foolish thing in the world, but to try to identify areas of vulnerability. [It’s] like saying a bridge is fragile. I can’t predict which truck is going to break it, so I have to look at it more in a structural form — what physicists call the percolation approach. You study the terrain. You don’t study the components. You see in finance, we study the random walk. Physicists study percolation. They study the terrain — not a drunk person walking around — but the evolution of the terrain itself. Everything is dynamic. That is percolation.
And then you learn not to try to predict which truck is going to break that bridge. But you just look at bridges and say, “Oh, this bridge doesn’t have a great foundation. This other one does. And this one needs to be reinforced.” We can do a lot with the notion of robustness.