What lies beyond AI efficiency

IT these days is AI this, AI that. Some shops have seen productivity jump several times over. Others tried it and, well, are still stuck at a slightly smarter search box. And some places have not even started — still at “AI? Is that something you eat?”

Ask the people for whom it is working, and they all say the same word. Efficiency.

So why chase it? To make room for the real work? To cut cost? There are probably lots of reasons. I still have not met anyone who can actually answer that.

I think efficiency only holds if something is already working. You take a mechanism that runs, make it smoother, raise the yield. That is efficiency. And yet, quite often, efficiency shows up first.

When you are facing a real business problem, you generally cannot be efficient. That is exactly where the value is. You shape it into a form you can repeat. Only then should efficiency appear. Efficiency itself is just something you are expected to do. It is not where you pull ahead.

I sometimes wonder what sits beyond efficiency. After an efficient stretch of work, what usually waits is the next piece of work. If that work is the real thing, fine. More often it is muddy, plain, and unchanging — which is not a complaint, by the way. You make that efficient too. More work arrives. It loops.

So when do you face the essential problem? Not in the spare hours after you have become efficient. I think you should face it now. Efficiency only exists on the far side of that. It looks like a difference of order. The difference is large.

So that's the thing!

What sits beyond efficiency? What is efficiency for? I expect I will keep asking people that.