If you want real emotion you would need to evolve it and so you would have to present the same kind of environment to the developing mechanism to make it develop these qualia we call emotions.
This doesn't make any sense. Are you proposing that qualia are transcendental?
He's saying (and I agree) that a programmed simulation is not the same thing as a true emotion.
It's kindof what i'm saying.
I'm not saying that a programmed simulation cannot do it in principle because i think it can.
What i'm saying is that it's a futile excersize because the easiest way to get human intelligence is to just use a human brain. A simulation like we are capable of running would not be sufficient to account for all dynamics in the brain. You will have to consider stuff like quantum mechanics that are simply available systems for exploitation by evolution.
I'm saying that you cannot define emotion by simply programming it as a chain of abstractions of the physical dynamics.
It proves to be too specific to capture in an algorithm.
You can generalize stuff, but in evolution details are just as important as the general structures.
Every single of the trillions of cells in a human body contain the whole DNA sequence. They all use bits of that code to interact with their environment. It's a mind-bogglingly specific system.
That is why people try to
grow complexity nowadays, instead of designing it. It turns out evolution is pretty efficient at this process despite its seeming inefficiency of exploring a possibility space.