The evolution one is one I've had a long-term passion for, going back, in some sense, to when I was an undergraduate. But yes, sometimes there are other subjects that I'm interested and if I had an extra lifetime, I would work on. One of them, for example, is in health. I mean, I could easily imagine I could have become a doctor, and what I'm really interested in the human body... everybody's focused on genetics, and I think that's like looking for the key under the lamppost, because genes are so easy to read and write. But the fact is, the difference between a sick person and a healthy person is almost never their genes, unless they were always sick. They pretty much have the same genes that they always had. It's in their body state. So if I take the computer analogy: when a computer is broken, it's almost never the binary of the computer, it's the runtime state of the computer that's broken. And our runtime state is not in our genes. It's in the... mostly in the proteins in our body. It's the things that are written by the genes. So our genes create proteins and those proteins turn genes on and off and they modify other proteins, and... so those are the things that really do things.

And we haven't paid much attention to them, because proteins are much harder to work with than genes. And they're also kind of analogue rather than digital. A gene really is a digital, literally four-letter code, whereas a protein, it matters how much there is of it. And so it's an analogue amount. And so one of the things that I've spent a lot of time doing in the last decade, is developing instruments that measure proteins floating around in the blood to great repeatability and accuracy. And this isn't a new idea, but it's an idea that's mostly been done by biologists, and biologists aren't really engineers, and so they tended to try to do these measurements in a laboratory by conventional techniques. And they... there are so many steps in the process that they tended to get very inconsistent results. So this field of proteomics got kind of a bad name, because people couldn't reproduce the results and the results had lots of noise in them and so on.