Hacker News new | ask | show | jobs
by yazanobeidi 2523 days ago
You should take another look at Eliasmith’s work. The neural engineering framework aims to be biologically plausible, with some assumptions made to simplify the math. Cellular processes are not the focus of the model. It however attempts to replicate these dynamics with simplified “components”. It’s one of the only, if not the only, neural models that is consistently supported by experimental data.

Their website is https://nengo.ai and I am not affiliated with them just an alum that took his course.

1 comments

I'm not against modeling aspects of the brain based on what we know, just don't call that "whole brain" simulation/model.
What would you call SimCity if not a city simulator?
It's a city simulator, not specific city simulator or even close. It is not accurate either as accuracy was not the main goal, but fun.

Human brain is a very specific and complex brain which we do not understand - vastly less than even cities at any scale. And to produce any useful answers, the model has to be accurate to some degree - and we don't know how much. HBP indeed had no question which it tried to answer - it's more like climbing a mountain because it exists.

I have no problem with calling SimCity a “whole city” simulator. Why? Because we have a good functional model of a city. We understand its operation on any level. Not the case with the brain.