Many researchers in cognitive neuroscience and deep learning seem to assume that inference in the brain is hierarchical.  Our new Nature Reviews Neuroscience paper questions this assumption and proposes the shallow brain hypothesis. #neuroskyence#brainwww.nature.com/articles/s41...

Have been thinking about this a lot. Such a good, important statement for AI aspirations. The review feels (necessarily and justifiably) "shallow" because each subcortical region has something very different to say about what "shallow" means--functionally and structurally.

When retrograde degeneration was discovered we thought of the brain in terms of thalamo-cortical paths. Then Kuypers and Pandya and separately Jones and Powell demonstrated a hierarchical cortico -cortical organisation. Somehow we need to keep BOTH types of organisation in mind.