Thought can move faster without necessarily moving deeper.

Performance may improve while something underneath changes.

The visible curve can keep rising while cognition takes a different path.

A while ago I wrote about what I called the velocity of thought. And now, I want to push on this idea a bit more, particularly in the context of AI's role in thinking—with us and for us. We've spent decades building machines that moved information faster, but large language models seemed to be doing something different. They appeared to influence the "movement of ideas" themselves rather than simply accelerating the transmission. Questions produced new questions, unexpected concepts connected, and thinking felt active in ways I had not experienced.

Over time, this picture became more complicated. I had called it the velocity of thought, the feeling that ideas themselves had gained a sort of forward movement. Later, I found myself writing about cognitive surrender, those moments when thinking is handed off rather than exercised, and anti-intelligence, the idea that fluent AI output can emerge without understanding itself being present. What emerged from this cognitive journey was the idea that thought can accelerate without becoming deeper and performance can improve while participation is reduced.

So I'm beginning to wonder whether velocity was measuring the wrong........

© Psychology Today