[russdill]

Just bring it down to the basics. Our brains operate on neurons, neurons operate on physics, physics can be fully simulated by computers.

This looping, CPU, "programming it in", and app concept are not the direction machine learning is going in. That's not how deep learning and neural networks work. You can integrate them with an app and do looping yes, but you can also just connect them to each other. No looping, no "programming it in", no apps.

[friendlybus]

Our brains don't solely work on neurons. There was a neuroscience video with 3-4 prominent people in the field dispensing of the idea that brains are computers. They squirt fluids around, it is an unknown. There are plenty of things about existence that physics does not capture.

Frankly I'm not saying ml is programmed in, only the initial conditions are, which is where the meaning is. We have hired a lot of low income earners to classify images for image recognition, which is the outsourcing of discerning meaning from the CPU to the he human. These kind of broad discussions don't go anywhere here, I should go somewhere more philosophical.

[gus_massa]

Fluids in the brain carry a few hormones from here to there and the distribution of neurotransmisors have an important role, but they act very slowly. If we ever have a 100% bug compatible model of the neurons of the brain, adding the flux in fluids will be easy. They can be modeled with a very low number of slow changing variables.

I don't expect to see ever a 100% bug compatible model of the brain. I expect to see some system that does a somewhat similar calculation and produce results that look similar to the result in the brain. Something like the eye and a camera.