Sunday, October 25, 2026

Pearls before AIs

There's an interesting phenomenon I've noticed and I'm not sure I have a good word for it.

One has various ideas jumping around in their head. Some of them are little pearls, beliefs big and small that seem significant and some of which may be quite old. They are all... cherished.

Many such ideas are about work so a programmer will have many software ideas. But you could have an idea about a great location for a new bridge in town or the idea that what your region really misses is some craft cider.

What AI really does well is implement many of these software ideas, quite decently. So you are put in the situation of seeing that precious nugget in your head materialize and finding it less impressive in reality.

The truth is that software is often much prettier when conceptualized than in reality.

But these ideas are important. It kinda hurts.

We can assume some implementations are really poor and thus do no justice to the great idea behind them. Maybe even manage to ruin the enthusiasm for that idea in the future. Which is a shame.

Though some implementations surely are decent.

I think some of these ideas were just supposed to be forever admired. When small, they are a sort of ego nugget. When big, they are a sort of muse, a guiding light.

"But I, being poor, have only my dreams;
I have spread my dreams under your feet;
Tread softly because you tread on my dreams." (Yeats)


No comments:

Pearls before AIs

There's an interesting phenomenon I've noticed and I'm not sure I have a good word for it. One has various ideas jumping around ...