Hacker News new | past | comments | ask | show | jobs | submit login

"consciousness" has a definition: You.

It's a common misconception that there's no way to detect consciousness. It has no qualities and so cannot be measured, but consciousness can be detected by other consciousness. Since you are conscious you can detect consciousness in other beings.

What's actually happening is that what we call consciousness is unitary and co-extensive with the entire Universe. When two people merge consciousness' they are just increasing the bandwidth between themselves until they are effectively a single organism, the separate self-models in each person merge into an integrated self-model. People do this spontaneously sometimes when the sex is really good.




I give you a straight forward notion that is scientific and concrete (the notion of "agent") - in other words - all that the notion of consciousness isn't, and you reply with new age hocus-pocus.

Of course you're free to think about the macrocosmic consciousness and the tantric web of relations that tie subject to object. But that isn't getting us any closer to understanding the topic. It's just a 1000 year old mostly forgotten philosophy. Even in tantra, it is considered that the subject acts through three energies or processes - iccha: will, jnana: knowledge and kriya: action - the same three aspects that define an agent.

The agent approach has game theory and reinforcement learning to back it, with real applications and results (such as Alpha Go). It does not presuppose a spirit, it has no mind-body problem, and is testable (no "first person" firewall around it). Given all that, I think the agent theory is much more parsimonious that the spiritual and philosophical treating of consciousness.


> I think the philosophy community (that was scared by behaviourism decades ago) needs to take a good hard look at reinforcement learning and representation learning. They can clarify many philosophical problems.

I agree with that.

> The hard-problem of consciousness was a detour, a wrong direction. It led to nothing in the end - just wasted decades. It was dualism in disguise.

I don't agree with that.

> You don't even need the notion of consciousness - just use the concept of agent, agent-environment relation and reinforcement - all being well defined, sound concepts unlike consciousness which doesn't even have an official definition.

It does have a definition: you.

What is the goal or purpose for which "You don't even need the notion of consciousness"?

> I think the agent theory is much more parsimonious that the spiritual and philosophical treating of consciousness.

To what end?


Minsky thought that consciousness is pretty much just a meaningless word, and that it's more interesting to divide it into pieces that you can work on: https://youtu.be/AO7F0n2Dclc?t=20m13s

These are nice lectures in general, too.


> Minsky thought that consciousness is pretty much just a meaningless word

In a sense, it certainly is [a meaningless word], in that it is an attempt to refer to something that has no qualities, so you can't really talk about it. If it weren't for the fact that you yourself are conscious there would be no way (or reason) to talk about it at all.

Now, I don't object to the idea that Minsky's work (or "agent theory") would usefully inform philosophy. In fact I agree with that whole-heartedly. The relatively unsophisticated epistemological stance of much of modern philosophy could be improved by a greater familiarity with computers et. al., in my (pretty ignorant actually) opinion.

What I objected to was the idea the the study of the so-called "Hard Problem of Consiousness" was "a detour". It can't be studied scientifically but it can still be studied.

Philosophy is the womb of science and the vestibule of mysticism.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: