Oh myrecent paper introducing some basic, fundamental problems confronting evolutionary theory. For whereas man-made machines may have a great number of components, such machines are specifically designed to limit the number of interactions. The components only interact with a small number of other components and a matrix describing these interactions would be very sparse. Not so for many biological systems. The paper shows that the magnitude of the interactome—the sum total of all interactions in systems such as the nervous system—is on the order of Bell’s number, which scales faster than exponentially. Indeed, for n discrete components, the logarithm of Bell’s number is n[log(n) – 1].
This means that even if we believe, with Moore, that compute power doubles every 18 months or so, it nonetheless will be impossible to analyze even a single synapse for another few millennia. And the visual cortex of a mouse? That will take about 10 million years to analyze.
But what if even Moore was pessimistic? Or what if only a tiny fraction of the interactions actually need to be characterized, say 0.001%? Such luck would not appreciably change the answer.
Technological advances and modeling simplifications do little in the face of the curse of dimensionality and researchers now use the term “complexity brake” to describe the resistance of biological systems to computer analysis. In fact, this brake is only going to become worse, for the more we learn about biological systems, the more complexity we discover:
Improved technologies for observing and probing biological systems has only led to discoveries of further levels of complexity that need to be dealt with. This process has not yet run its course. We are far away from understanding cell biology, genomes, or brains, and turning this understanding into practical knowledge. The complexity break is very apparent in the figure (shown above).
All of this not only means that the analysis of such biological systems is impossible, it also means that the evolution of such biological systems is impossible, or at least scientifically unlikely. For the problems in analyzing these systems also apply to evolving these systems.
Not only is there a large number of interactions but, unlike a bottle filled with a gas which can be characterized by averages and distributions, in biological systems the specific, individual interactions matter.
It would be serendipity on steroids to say that evolution, with its limited experimental powers, designed and created a few basic components which then, as luck would have it, combined in such a way to produce far greater complexity and emergent behaviors.
And that, here in the twenty first century, is increasingly what evolutionary theory is all about: serendipity.