This is about pilots. The kind of pilots you do before spending a lot of money on something. The idea of them bothers me.

I don’t want to be got wrong. Pilots that don’t make presumptions about what comes next - ones that are genuinely about learning - are fine. It’s those where the logic is:

"We have a complex problem that, of course, needs a complex solution. But it's a big commitment and, naturally, with ambition comes risk. So, as part of the solution, we will do a pilot..."

Pilots that are about reducing the risk of a particular solution, skew (or more likely preclude) learning about other possible solutions. They are often based on the presumption that the complex design - done by clever people - is essentially right. Which is often wrong.

In complex situations it's better to do experiments: to start with a hypothesis rather than an answer.

And it's better to call them experiments; so people involved get the right idea.

As Joan Armatrading said (about those on target and whose aim is straight) ...