Consciousness is not an impossible problem. It only becomes absurd if it's tinted of mysticism and dipped into a half-digested understanding of the current scientific consensus.
First, let's demystify conscience. Scientifically, there's no soul - it's a religious concept that has no meaning outside of it. What's left of conscience is its shell, its interface.
If we can model something that behaves like conscience, well, we created a genuine conscience.
There are two sides to the shell of conscience. One is the outer shell: the problem is to build something that appears to be conscient. This is a reasonably hard CS or neurobiology problem but by no means impossible. We are getting reasonably close to generating seemingly conscious automata. Surely it's possible to see specific parts of the brain connected to this function: for example, the speech center, and so on. So: hard to do, but doable.
The other part is the inner shell, or "self-conscience." This has to do with perception, mostly, and abstraction. We have already created expert systems that can perceive and give meaning to a lot of inputs.
The trick is creating a system that can perceive itself, its state like it sees the world. This will require a general AI or strong AI, which is currently believed by many experts to be possible although super hard to do. Again though, this is a CS problem, it's addressable by science, and I have no doubt it's a matter of time until we can settle it.
A machine with both characteristics would be just as alive as you or me. It would perceive the world semantically. It would understand to be conscious, and we would recognize it to be conscious too.
At that point, the question of consciousness will become once more only attractive to philosophers and priests. The rest of us will have less problem accepting one of these machines as "alive and conscious."
Consciousness is not an impossible problem. It only becomes absurd if it's tinted of mysticism and dipped into a half-digested understanding of the current scientific consensus.
First, let's demystify conscience. Scientifically, there's no soul - it's a religious concept that has no meaning outside of it. What's left of conscience is its shell, its interface.
If we can model something that behaves like conscience, well, we created a genuine conscience.
There are two sides to the shell of conscience. One is the outer shell: the problem is to build something that appears to be conscient. This is a reasonably hard CS or neurobiology problem but by no means impossible. We are getting reasonably close to generating seemingly conscious automata. Surely it's possible to see specific parts of the brain connected to this function: for example, the speech center, and so on. So: hard to do, but doable.
The other part is the inner shell, or "self-conscience." This has to do with perception, mostly, and abstraction. We have already created expert systems that can perceive and give meaning to a lot of inputs.
The trick is creating a system that can perceive itself, its state like it sees the world. This will require a general AI or strong AI, which is currently believed by many experts to be possible although super hard to do. Again though, this is a CS problem, it's addressable by science, and I have no doubt it's a matter of time until we can settle it.
A machine with both characteristics would be just as alive as you or me. It would perceive the world semantically. It would understand to be conscious, and we would recognize it to be conscious too.
At that point, the question of consciousness will become once more only attractive to philosophers and priests. The rest of us will have less problem accepting one of these machines as "alive and conscious."