The room is not just an input-output system. It also has to have state. Without state the whole argument falls apart. > But the notion is not far from the bon

The room is not just an input-output system. It also has to have state. Without state the whole argument falls apart. > But the notion is not far from the bone-headed simplicity of early AI and machine translation efforts. As someone who studied Machine Learning 20 years ago, could you be a bit more respectful please? First, pretty much nobody (working on serious ML) in that era thought they were going to build anything actually intelligent, conscious, or thinking. Second, in this supposed era of boneheaded simplicity, a LOT of foundational stuff we use in today's AI was discovered and written. I'm not sure how familiar you are with AI history, to name an example, we already were working on neural nets and discovered error backpropagation in the 70s/80s. Then came a long period during which neural nets somehow didn't really perform very well, and ML was looking at completely different types of classification algorithms (what's happening with SVMs today?). If we had stopped there, you might have lumped in the early perceptron with the "boneheaded simplicity". But instead we developed batch normalisation and a bunch of other techniques, and now neural nets are state of the art.