We’ve stepped or perhaps even fallen into a world where machines speak with seductive confidence. And today, they offer answers that feel polished and curiously satisfying. But I've begun to think not only about what that does "for" us and more about what that can do "to" us.

Large language models are now today's fluent performers, even digital actors, that articulate ideas with rhythm and poise. They don't even give us time to doubt, only to acquiesce to the brilliance that is just about instantaneously generated. But beneath that smooth surface, something altogether different resides. They don’t really know, they just compute.

The simple (and complex) truth is that LLMs arrange words in the order most statistically likely to appear next. The magic is without intrinsic memory, belief, or any genuine sense of the world they’re describing. In an earlier post, I called this anti-intelligence—not as an insult, but as a way of naming its "inversion" of human cognition. AI isn’t a lesser form of intelligence—it's something structurally different and antithetical to the human mind. And yet, despite that empty core, it produces something that feels uncannily like knowledge.

But anti-intelligence doesn’t stand alone. It has a companion........

© Psychology Today