Programming the Human Mind:

[I_am_not_mafia]

I don't think it really is much about the programming language, but more about the the hardware itself... The "computer" that would one day run a human brain would not be a general purpose computer like the one we are using now. I think it has to be a specially made computer designed specifically to emulate a human brain. It might physically look like a human brain or at least have a physical representation of neurons in its circuits.

If you're planning on simulating or emulating the human brain then we're using specialised hardware right now in the form of NVidia graphics cards. It's still not enough. Although massively more parallel than a normal CPU, we need hardware with many, many orders of magnitude more connectivity between the processors.

And I am not sure if a programming language can accurately represent the human brain. It's much more complicated than hungry->seek food. What if you are hungry and at the same time you were performing an important task, you would not seek food then. Or if your life was in danger, food would be the last of your worries.

One way of doing that is to emulate the effect of neuro-modulators on neurons. They have long lasting effects that either excite or inhibit layers of a neural network. They act as non-linear gain controls and allow a self organising system to either settle into or be driven out of stable states. But what you then effectively have is not a computer program so much as a simulation of a physical system.

We know that you can have intelligence using neurons, we're each examples of that. But you can simulate a neural network on computers. I've also figured out how to simulate my own neural networks in a garden using pipes and cisterns that gate and release water when the water level reaches a threshold. It's just me as an observer deciding that both these are examples of neural networks. What they actually are, are self organising systems that perform certain functions. And if a system, whether implemented on a computer, in a garden using pipes and cisterns or using biological material performs can perform the same functions as the brain then why should we have to assume that intelligence has be in the form of neural networks?

We might end up using an entirely different model for creating artificial intelligence like dynamical systems for example. They will perform the same functions but be better suited to the underlying architecture they are implemented on.

Also what do you interpret as being hungry? Having a headache? Stomach churning? Not having eaten in a while? How would a cpu understand hunger? It can't feel hunger, it has to simulate it, and when should it simulate the hunger? Also not all hunger are equal, sometimes you are really hungry like you might die, and sometimes you just crave something.

Exactly. The computer needs to be situated within a body. It needs it's own version of hunger. In the same way that we do not experience the same kind of emotions as say a dog or a bird. A vacuum cleaning robot won't ever feel hunger in the same way that we do but then it will have a battery level that needs to be maintained. If this drops from 70% to 69% then no problem. If it drops from 4% to 3% then this is major cause for alarm. It also needs to stay safe and has to be fearful of dangerous situations or anything it does not recognise. And as is the case for every other agent with emotions it has competing needs but only one body.

We are each a product of our brain, body and environment.

All the above and we only scratched the surface of one topic: hunger. I heard there is a human brain project in europe. I am looking forward to seeing if it goes anywhere, however it has many skeptics, me among them.

I don't think anyone is expecting to create human level intelligence from it. But the project will definitely give us some useful information and experience about how the brain works.