3 ms·
The Consciousness Deniers
- visarga 9y ago> to reduce consciousness to behavior and dispositions to behavior is to eliminate it. To say that consciousness is really nothing more than (dispositions to) behavior is to say that it doesn’t exist I beg to differ. Consciousness is a vague, "suitcase word". It is close to intuition, but resists being put into words or being explained in a scientific framework. But what is concrete and observable is behaviour, environment and reward signals. They are the basis of the constant environment-adaptation (homeostasis) that we call consciousness. I would not deny consciousness, just put it aside as a vague concept and replace it with Reinforcement learning. RL has had great results in AI recently (AlphaGo) and can explain everything including qualia. In my view, qualia is the evaluation of the current state with regard to future rewards (how good a state is for the agent). Qualia is necessary in order to select appropriate actions depending on the situation. It's not just a "what it feels like" first person thing. My critique to those who support the views of this article is that the concepts of "qualia" and "consciousness" are stuck and lead to no progress. It has been 2000 years since philosophy has been grappling with this problem using introspection and pure thinking. We need to go back to the data and observe environment, actions, values and rewards. These concepts are much better, they are concrete and implementable in software. They are much better intuition pumps that explain human, animal and AI behaviour. I think trying to create a separation between behaviour and consciousness is a wrong path to take. RL puts much more emphasis on the role of the environment in the development of concepts and intelligence. The environment is often overlooked by philosophers who look for neural correlates of consciousness in the brain. In fact the brain is a distillation of the environment, a model that depends a lot on the input data it learns from. Consciousness cannot be understood outside of its environment, it makes no sense outside, it's always consciousness of something, not consciousness in itself. In the end, it's all a game of life and death, a game of self replication and competition played by consciousness, the genes and the memes. What philosophers need, in my opinion, is to take a look at RL, deep learning, game theory and complexity theory. They explain how everything comes together in a more rigorous and parsimonious way.
- dekhn 9y agoNonscientific philosophers should probably stay out of the business of neuroscience.