And they are generated by their host piece by piece (brain growth). And they last longer than AI processors with basic maintenance anyone can do (eat/sleep). And they usually continue working when their programs break (eg hardware failures or gamma rays) while NN’s and GA’s don’t.

There’s hardly a comparison. If they want to, I’d like to see them start with using dynamic chunking instead of tokenization like the mind does. Also, they can tell me what pre-training method uses 20W per second for 1+ trillion parameters. I could use that right now.