Why a 'Language Model' (such as ChatGPT) is not even nearly AI

Here is an interesting alternative to the Artificial Neural Nets (ANNs) that are used with ChatGPT that models more closely how humans reason. It uses mathematical operations on 10,000 bit vectors.

https://www.quantamagazine.org/a-new-approach-to-computation-reimagines-artificial-intelligence-20230413/

"10,000 bit vectors" is merely numerical range to allow for higher resolution - of what?

The question at hand is on a principle level:

What could be defined as "Understanding" in reference to a machine, a man-made mechanism which can "Understand".

And even then, "Understanding" (or "Deduction" if you like) is barely the tip of the iceberg of traits required towards an ideal called "AI"