I know, I know. Molecular biologists are an easy target for us field types. Just the other day Dr. Corrie Moreau, ant systematist extraordinaire, repeated an old joke with gusto over a Friday Skype:

So a stranger comes up to a sheepherder and says, “If I tell you how many sheep are in your flock, can I keep one?” After a moment, the sheepherder nods his assent. “You have 634 sheep.” the stranger says. Awestruck, the sheepherder says, “You win, take your pick.”. As the stranger walks away with his prize under his arm, the sheepherder yells out “Hey, you wouldn’t happen to be a molecular biologist would you?”. Now it was the stranger’s turn to be surprised. “How did you know?” he asked.  “You’ve got my dog.” the sheepherder replies.

So, anyway.

I’m perusing a recent issue of Science when I came across an article on how DARPA, a research arm of the defense department, is trying to automate the process of coming up with promising hypotheses. To which I say, “Hey, anything that helps, more power to ya.” But it was the target of their first efforts that floored me.  I’ll let Dr. Larry Hunter explain.

Building a system that actually produces scientific insight will not be easy, says computational biologist Larry Hunter of Smart Information Flow Technologies in Minneapolis, Minnesota, a co–principal investigator of one of the teams. The artificial intelligence community doesn’t have a strong track record at building systems that can develop useful causal hypotheses, he says. But molecular biology is a good place to try, he says, because it’s an area in which common sense plays a minor role; most of the knowledge is technical and available in textbooks and papers.

Now, I know, I know. Different challenges for different disciplines. Some of my best friends are molecular biologists. Just don’t let them near your dog.

From Science Magazine  30 January 2015 347 (6221): 465