I think characterizing this discussion as being about whether zombies are conceivable, as Eliezer does here, prevents productive discussion. That is not the issue, and Eliezer basically admits that in the last paragraph. Of course they are conceivable. We all know what we are talking about here.
Eliezer’s basic argument is that zombies are impossible, not that they are inconceivable. And I agree that they are impossible. But the fact that he has misrepresented the nature of the argument makes it difficult to have a productive discussion of the issue.
Suppose we have a grid of pixels, with pixel #1 located at position 2,1 ; pixel #2 located at position 2,2; pixel #3 located at position 2,3; pixel #4 located at position 2,4; and pixel #5 located at position 2,5.
The pixels are in a straight line on the grid. Now suppose someone says, “Could there be a series of pixels, all in exactly those positions mentioned, but in such a way that the pixels are not in straight line?”
In this case, asking whether or not the situation is conceivable is not a helpful question here. But we do know that the situation described cannot happen. I would say that zombies are essentially the same situation—something physically identical to a human is a human, and has all human properties, including the property of consciousness.
One difference though is this: we think we understand why the pixels must be in a straight line, but we do not think we know enough about the physical properties of humans to say why they must be conscious. We just know that humans are in fact conscious, and this is enough to tell us that zombie humans cannot actually happen.
Given the terms that physics usually uses, in fact, a deductive argument to the conclusion, “humans are conscious,” is impossible, since “conscious” is not one of those terms. But in the same way, if we start from premises that only say things about the positions of individual pixels, and nothing else, we cannot formulate a deductive argument that the pixels must be in a straight line. That does not mean that the pixels might fail to be in a straight line, nor does it imply that a human could fail to be conscious. It simply means that our account, whether the physical account of the human, or the one about the positions of the pixels, is an incomplete account of reality.
And I suspect that this last point is my real disagreement with Eliezer. I think he believes that the physical account is a complete account, and likewise that an account of the pixels including nothing but the individual positions is a complete account of the pixels. If so, I think he would be mistaken in both cases.
I think characterizing this discussion as being about whether zombies are conceivable, as Eliezer does here, prevents productive discussion. That is not the issue, and Eliezer basically admits that in the last paragraph. Of course they are conceivable. We all know what we are talking about here.
Eliezer’s basic argument is that zombies are impossible, not that they are inconceivable. And I agree that they are impossible. But the fact that he has misrepresented the nature of the argument makes it difficult to have a productive discussion of the issue.
Suppose we have a grid of pixels, with pixel #1 located at position 2,1 ; pixel #2 located at position 2,2; pixel #3 located at position 2,3; pixel #4 located at position 2,4; and pixel #5 located at position 2,5.
The pixels are in a straight line on the grid. Now suppose someone says, “Could there be a series of pixels, all in exactly those positions mentioned, but in such a way that the pixels are not in straight line?”
In this case, asking whether or not the situation is conceivable is not a helpful question here. But we do know that the situation described cannot happen. I would say that zombies are essentially the same situation—something physically identical to a human is a human, and has all human properties, including the property of consciousness.
One difference though is this: we think we understand why the pixels must be in a straight line, but we do not think we know enough about the physical properties of humans to say why they must be conscious. We just know that humans are in fact conscious, and this is enough to tell us that zombie humans cannot actually happen.
Given the terms that physics usually uses, in fact, a deductive argument to the conclusion, “humans are conscious,” is impossible, since “conscious” is not one of those terms. But in the same way, if we start from premises that only say things about the positions of individual pixels, and nothing else, we cannot formulate a deductive argument that the pixels must be in a straight line. That does not mean that the pixels might fail to be in a straight line, nor does it imply that a human could fail to be conscious. It simply means that our account, whether the physical account of the human, or the one about the positions of the pixels, is an incomplete account of reality.
And I suspect that this last point is my real disagreement with Eliezer. I think he believes that the physical account is a complete account, and likewise that an account of the pixels including nothing but the individual positions is a complete account of the pixels. If so, I think he would be mistaken in both cases.