4 ms·
That’s a plausible explanation for a necessary condition for subjective experience, but it’s diffult for me to accept it as sufficient. The mechanism - like int
by paraschopra 8y ago
That’s a plausible explanation for a necessary condition for subjective experience, but it’s diffult for me to accept it as sufficient. The mechanism - like integrated information theory - could account for the easy problem of consciousness but not the hard problem of consciousness.
A scientific theory of consciousness would make new predictions that can be tested. For example, if we take self-encoding to be actually representative of subjective experience, can we find a novel encoding that when triggered in a subject makes them report a sensation that they’ve never experienced before? That’s how new scientific theories that propose new entities like quantum ‘spin’ are tested. New fundamental entities must have new testable predictions otherwise they’re just-so stories.
That said, the easy problem is a path towards the hard one.
- visarga 8y ago> A scientific theory of consciousness would make new predictions that can be tested. The theory Schmidthuber mentioned is called Reinforcement Learning, a field of AI. RL also happens in biological agents, and we have known about it since Pavlov. A 'prediction' of RL is AlphaZero, the agent that beat humans at Go. Using the theories of RL we managed to accomplish what seemed impossible, and it was not just brute force, it was based on an artificial evolutionary program of agents battling each other, using neural networks to learn from past experience. It's all captured in a nice formula: https://en.wikipedia.org/wiki/Bellman_equation https://en.wikipedia.org/wiki/Bellman_equation