The Illusion of Thinking

> crucial questions about their true reasoning capabilities

It may be found once and for all, these are automated statistics of thoughts presented beforehand.

Those ideas and circumstance diversely represent in input sets will influence accuracy of output.

That is to say, it is only doing what it has learned from us to degrees of ever more complexity (more indexes).

The LLM is an index with vector tables computed in real (answer) time.

These are not inventing new views they are repostulating existing experiences into a situational context. They are stitching together a narrative, that’s it.

Btw, I love the think feature. “Tell me something profound” was an adventure of tangential consequence.