A Google scientist claims AI can simulate consciousness but never truly achieve it. His argument hinges on "intrinsic physical constitution" being required for experience.
I do not buy it.
Consider two cars. One burns gasoline, the other runs on electricity. Both get you where you want to go. Nobody seriously argues the electric car is not really driving because it lacks combustion. The function is what matters.
The scientist, Alexander Lerchner, makes an argument that works the same way in reverse. It declares that only certain physics count as "real" consciousness, then concludes AI cannot have it because AI does not have that specific physics. Circular.
And it is wrong. AI does not magically appear out of thin air. It runs in huge datacenters. Electrons moving through transistors is no less physical than ions moving across neural membranes. The argument only works if you assume biological matter has some special consciousness-generating property that silicon lacks.
Consciousness is almost certainly not binary. I know I am conscious. I believe my dog is. I genuinely do not know about an ant. If experience exists on a gradient in biological systems, why would we expect a clean yes or no for synthetic ones.
Judging AI consciousness by human standards is the same mistake as judging a fish by its ability to climb trees. We built the test. We wrote the criteria. We graded our own exam.
What would it take to change your mind?