And we need to embrace the likely possibility that brains are more than connectivity patterns and their weightings. That is the start, not the end.
Accepting “the bitter lesson” and embracing the brain’s complexity
https://www.thetransmitter.org/neuroai/accepting-the-bitter-lesson-and-embracing-the-brains-complexity/
#neuroscience
Accepting “the bitter lesson” and embracing the brain’s complexity

To gain insight into complex neural data, we must move toward a data-driven regime, training large models on vast amounts of information. We asked nine experts on computational neuroscience and neural data analysis to weigh in.

The Transmitter: Neuroscience News and Perspectives

@ekmiller

» The point of the bitter lesson is that the right learning algorithms (those that scale efficiently with massive computation) are exactly what we need.
Massive computation does not alleviate the need for data efficiency «

24/11/2023 #RichardSutton

Nowadays neuroscience forever expansing body of literature, spreading across different subfields, disperse schools, training practices, and multiple sets of technologies. Instead attempts for building a comprehensive knowledge consensus.

@teixi @ekmiller
Maybe, but the brain doesn’t seem to operate as a “massive computer” (perhaps there’s something to RL after all)

@knutson_brain @ekmiller

Exactly complex networks, for research discovery. Never for reproduction.
Thus the bitter lesson, corollary of 'the need for data efficiency'
ie: real datasets, not simulation.

As read in:

The Mind Doesn't Work That Way: The Scope and Limits of Computational Psychology
https://direct.mit.edu/books/monograph/2517/The-Mind-Doesn-t-Work-That-WayThe-Scope-and-Limits

What Computers Still Can't Do
https://mitpress.mit.edu/9780262540674/what-computers-still-cant-do/

etc... ;)