David Chalmers: Could a Large Language Model Be Conscious?