Nobody wants to hear about your biologically plausible computational framework

I go back and forth with this guy on twitter arguing about HTM vs Bayesian active inference. I’m pretty happy with this “internet own”, in response to him saying “Nobody wants to hear about someone else’s computational framework, especially a biologically plausible one. :-)”

Anyway you folks are wonderful. Anyone reading this forum genuinely looking for biological solutions to intelligence will find a wealth of ideas and potential avenues forward.

Thanks for taking an interest in biologically-inspired intelligence frameworks and being large and smart enough to help me perform this zinger. :stuck_out_tongue:


Chris, if you are reading this, no hard feelings. :lollipop:

I always seek out dissenting views on ideas or theories I’m excited about, as I’d rather consider them up-front than encounter them later on. I also really appreciate that I can find open discussion on alternative theories on the forums here, right alongside (and in contrast to) HTM.

In the case of general public commentary on Numenta’s research, I find it funny that two of its biggest detracting arguments are almost contradictory:

  1. It’s too ambitious - the brain is too complex to try to model holistically with a unified theory
  2. It’s not ambitious enough - we don’t want to be limited by a biologically constrained model, we can design an intelligence so much better than nature can