And one of the things I realised is that the problem is very similar to the problem of making a semiconductor line, which is... I remember the days when you could make chips and you could make, sort of, small chips on a lab bench, but as they got bigger and bigger, you had trouble reproducing them and you could only make chips with a few transistors on them. Making microprocessors that had millions of transistors on them required very carefully automating the line and then going through and figuring out all the sources of variation, engineering them out one at a time. It was a big engineering job that always required just pushing down the sources of variation. So we did that engineering job at Applied Proteomics and really pushed down all the sources of variation so that we could put a drop of blood in and get out hundreds of thousands of protein markers, and... but if we put that same drop of blood in or another drop of blood from the same sample, the next day we would get exactly the same answer out. And that had never been done before. So... at least to that level of precision.

So that gives us a view into the body that ought to show what's going on before you start showing symptoms when you get sick, for example. So when you get sick, it's not usually when something went wrong. It's when something went so wrong that your body can't hide it. So usually the sequence is: your body is healthy, something starts to go wrong, but your adaptive systems manage to mask what's going on. Sometimes they manage to just fix it and you never know you were sick. But sometimes that gets out of control so far that your body can't mask it and you start showing symptoms. That's usually when people go to the doctor. But by the time they go to the doctor, it's already way into the process and the doctor has to deal with all these secondary problems that are happening, not just the original cause.

And so doctors treat sick people. Or doctors do a little bit of saying: don't smoke, don't overeat, exercise, never get sick. But actually, the interesting stuff is in between those. The interesting way for doctors to work would be if they could see that you were on your way to getting sick, things are starting to go wrong, but you pre-empt the sickness. The only way to see that would be to measure what's going on inside your body before you start showing the symptoms. So I think this has tremendous potential for completely changing the way that medicine works. If we could constantly monitor what's going on in your body, notice when something is starting to go wrong, help the body fight it off before you ever get sick, that would be fantastic. We could... and I have simple examples where we can do that. The first example is I can see from a drop of blood when you are starting to grow a colon polyp. And a colon polyp is a precursor to colon cancer. So if you know you're starting a colon polyp then you can go in with a colonoscopy, remove the colon polyp, it'll never develop any cancer. So there's no reason anybody should die of colon cancer, if you can see colon polyps growing. But right now, the only way to do that is with a colonoscopy, which is not that accurate and it's kind of dangerous and so you can't do it very often. And so even people that are high risk don't get colonoscopies very often. But if you had this blood test, they could get... you know, once a week, they could test their blood. And you could head off the disease before it ever happened. So there's a very specific example. I think there's going to be lots more examples like that. So I could imagine if I had an extra lifetime, just concentrating on that way of changing medicine. And part of me is tempted to do that, but you know, I think there are other people that will do that.