Also, isn t the whole project making some completely wrong assumptions? I heard about the ideas that neurons don t make synapses on their own and that astrocytes instead of just being support cells is acting like sculptors on their sculptures with research having focused on neurons mainly because eeg detectability. And to support this, that it is the underlying reasons different species with similar numbers neurons shows smaller or larger connectomes and they are research that claimed to have improved the number of synapses per neurons and memorisations capabilties of rodents (compared to those without) by introducing genes controlling the astrocytes of primates (thus I recognise this theory is left uninvestigated for protostomes and their neuroglia instead of the full fledged astroglia of vertebrates).
Of course, this would had even more difficulty to the project https://www.frontiersin.org/articles/10.3389/fcell.2022.931311/full.
Having results with completely wrong assumptions doesn t means it doesn t works. For example, geocentric models were good enough to predict the position of planets like Jupiter during the medieval time but later inadequate and hence the need to shift to simpler heliocentric models. Getting all clinical trials on Alzheimer of the past 25 years failing or performing poorly in humans might suggest we are completely wrong on the inner workings of brains somewere.
As an undergraduate student, please correct me if I said garbage.
For point 2, is it possible to use the system to make advance in computer ai through studying the impact of large modifications of the connectome or the synapses in silicon instead of in vivo for getting eeg equivalent? Of course, I understand the system might have to virtually sleep from time to time unlike the pure mathematical matrix based probability current systems.
This would be the matter of making the simulation more debuggable instead of only being able to study muscles according to input (senses).