Pistols

When you coach little kids in basketball, you spend most of your time teaching them not to watch the ball.

Watching the ball is what every nine-year-old wants to do. The ball is constantly moving and it’s where the action is. So they stand on defense and turn their whole body toward whoever has it, and the kid they are supposed to be guarding cuts to the hoop behind their back, and the coach yells, and the kid usually doesn’t understand what happened.

The drill I taught them was pistols. Both hands up at shoulder height, thumbs up, one index finger pointing at the man you are guarding and one index finger pointing at the ball (resembling pistols). Both arms up pointing at two targets, head on a swivel between them. You cannot lock onto either one. The second you do, the play happens in the place you are not looking. Me yelling “PISTOLS!” still haunts my dreams.

I’ve been thinking about this a lot these last few weeks.

The swivel

I’ve been a founder for over a decade now. The pace at which AI is moving right now I’ve got no comparison for. There’s something new every week. Often something new every day. A model, a paper, a product, a new architecture that didn’t exist on Monday and is reshaping how we’re building by Friday.

The job is not to keep up with all of it. Nobody can. The job is to develop a discipline for what you swivel toward and what you let go.

That’s what ‘pistols’ is. One finger has to point at the ball, every model release, every paper, every Twitter/LinkedIn thread that might be hype or might be the thing that reorganizes your category in six months. The other finger has to point at your man, the actual product you are building and the problem you got out of bed to solve. If you watch only the ball, you forget what you’re trying to do. If you watch only your man, the play happens in the place you’re not looking and you don’t realize it for nine months, and by then it is too late.

Head on a swivel. The whole time.

The truth

I am not the one with the hardest job in the room.

Most of my conversations are with Presidents, Provosts, VPs of Enrollment, CIOs. People who run institutions that are two hundred years old, that have boards and faculty and accreditors and donors, and that suddenly feel the ground shifting underneath their feet in real time. They are being asked to make decisions about AI with limited information, on a timeline they did not choose, against competitors that did not exist eighteen months ago.

The version of pistols I run is hard. The version they are running is much harder.

A founder gets to be wrong. Startups are built to fail fast and adjust. They don’t have that luxury. A VPEM who waits too long and a VPEM who moves too fast can both end up running an operation that’s no longer competitive, and they won’t know which one they were for two or three years. By the time the data’s in, the call’s already been made.

What this means

The mistake right now is not moving too fast or too slow. It is indecision.

AI isn’t going away. The institutions that move first will build the muscle, lower their cost to enroll, and compound the advantage. The ones that wait will fall behind, just slowly enough that it’s hard to see in the moment.

If I were sitting in one of these roles, I’d do three things immediately. Find the people already doing this well and learn from them. Look for real case studies with real numbers, not promises. Put AI into production in controlled ways and build the muscle now, before you need it.

Because the worst outcome in this moment is not being wrong.

It’s doing nothing.

That’s ‘pistols’ in practice.

We’re inside one of the biggest technological shifts any of us will ever see. Hands up. Head on a swivel. All of us.