Status Updates From How to Create a Mind: The S...

"We are apparently very eager to explain and rationalize our actions, even when we didn't actually make the decisions that led to them. So just how responsible are we for our decisions?"

Just realized this book was written in 2011...

Hierarchical Hidden Markov Models seem very similar to language transformers. What are the differences between these two technologies? Are transformers a subset of HHMMs, if so why did it take so long for people to build transformers when HHMMs have already been discovered for quite a long time?

"We need to factor the phenomenon of redundancy." I don't particularly understand why it is that we have redundancy in pattern recognition, especially if we accept the proposition that memory is hierarchical. If memory is hierarchical doesn't that mean patterns are simply instantiations of these hierarchies? So why would there be redundancy? Wouldn't hierarchies result in simplification?