The same way the Tononi’s IIT is testable (and false): it predicts that a Vandermonde matrix multiplier is conscious, and more so that you and I, if the matrix is large enough.
Nobody knows yet. That’s what makes it the Pretty-Hard Problem. Our ignorance of how to test it should not be projected onto the world and mistaken for proof that it is untestable.
It’s not merely the “pretty hard problem”. It’s the “impossible to attack, by definition” problem. If you use “consciousness” as a synonym for something such as “intelligence” then it becomes tractable, at least in principle, but you will always have those who insist you’ve simply changed the subject (including me).
Our ignorance of how to test for something we insist exists is not proof of anything, but it is strong evidence of a conceptual muddle.
Our ignorance of how to test for something we insist exists is not proof of anything, but it is strong evidence of a conceptual muddle.
In the case of consciousness, we do not have to insist on anything. We can simply point to our internal experience, and say, “this is what we mean, when we talk about consciousness.” That we have no explanation for how there could possibly be any such thing does not invalidate the experience, for even a faultily conceptualised experience is still an experience. No matter how the experience is reinterpreted, it obstinately remains an experience.
The conceptual muddle is in thinking that because we do not understand a thing, it therefore does not exist.
He is asking a question whose answer would be testable, so no “assertions about magic”. That’s the best one can do.
How does one test a machine for consciousness?
The same way the Tononi’s IIT is testable (and false): it predicts that a Vandermonde matrix multiplier is conscious, and more so that you and I, if the matrix is large enough.
Where can I find the experiments that tested the Vandermonde matrix multiplier for consciousness?
Nobody knows yet. That’s what makes it the Pretty-Hard Problem. Our ignorance of how to test it should not be projected onto the world and mistaken for proof that it is untestable.
It’s not merely the “pretty hard problem”. It’s the “impossible to attack, by definition” problem. If you use “consciousness” as a synonym for something such as “intelligence” then it becomes tractable, at least in principle, but you will always have those who insist you’ve simply changed the subject (including me).
Our ignorance of how to test for something we insist exists is not proof of anything, but it is strong evidence of a conceptual muddle.
In the case of consciousness, we do not have to insist on anything. We can simply point to our internal experience, and say, “this is what we mean, when we talk about consciousness.” That we have no explanation for how there could possibly be any such thing does not invalidate the experience, for even a faultily conceptualised experience is still an experience. No matter how the experience is reinterpreted, it obstinately remains an experience.
The conceptual muddle is in thinking that because we do not understand a thing, it therefore does not exist.
Descartes said all that in three words.
The conceptual muddle is in thinking that because a thing exists, we must be capable of understanding it!