Jump to content
  • Advertisement

Journal

Sign in to follow this  
  • entries
    122
  • comments
    246
  • views
    90766

A tentative Probe

Sign in to follow this  
Daerax

151 views

Suppose that the brain was completely deterministic in the classical sense, that is, we do not take quantum mechanics into account and imagine the brain to behave in a clockwork manner.

Then the concept of volitional movement should not be possible since in a clockwork object, everything moves, dependent on each other and nothing changes. It seems that the brain is a system of cells which can direct its own behaviour and change its own state. The behaviour of the brain is not the same one repeated over and over again, there exists ordered patterns and groupings and more importantly, self modification that have reason (to us) behind them. We define reason based on the behaviour of the brain which allows us to define reason and then say there seems to be reason in the brain's action. But is there actually reason to the brain's self modification? Any motivation? Why should such exist?

How is this possible? What is an AI expert take on this?

Suppose the brain was deterministic in the sense than given the same set of inputs (environment) each time it would output the same result. That is, the same morphisms or change of configurations will always occur due to the same actions being performed resulting in a complex change in mental, memory and emotional state etc. Which make it seem that behaviour was being directed when instead a new state is triggered where one imagines that such things as where to place some limb, some extra attributions to an internal representation of some object or what object is currently of interest are actually what occur. But then where does thought fit in? What processes initiates a state of sentience or even self awareness? What are the current means of begining to approach this in AI?

I will offer my non AI viewpoint that the key lies in language. I feel that the essence of a weak form of the Sapir Whorf Hypothesis must be correct. Indeed it supports my view that physics and any intepritation of reality is entirely subjective. For example, the article talks about the Native American Hopi, whose language reflects a thought process where space and time have no true meaning but rather, processes are key - for them Relativity would be entirely intuitive! That statement is so important because modern physics has begun to drop such things as space, time and motion. But i digress.

I instead think languages are a reflection of the thought process, the two are related although i feel thought determines language which in turn affects the form of the thought processes. I.e. culture. I conjecture then that the first step to an approximation of AI would have some formal system as the background to its operation. That is, place neural nets on a formal system. It operation would be built and categorized by its logic ang language. Instead of the *from logic* approach what about an *of logic* approach?

What happens when you have some state which you wish realized by you NN and have a language whose form is semirandom and evolve (using concepts from self organization in complex systems and GA's) towards that state based on the form of said language. Is this sensible Nick? Please correct my misassumptions.
Sign in to follow this  


1 Comment


Recommended Comments

Hmmm.. well, just off the top of my head (the part of my head where the conic section yields a point), I would say that NN's could simulate some aspects of languange. For example, I could go outside, examine the sky, and say, "My, the sky is blue", then I could train a simple feedforward backprop NN with environmental inputs and have it point to a distribution of responses similiar to mine. Then I could model the random part of the response with another neural net that approximates a probability distribution over several equivalent responses. Now, I would assume that this "brain" learns and forgets so that the distribution has a tendancy to go flat and also give bumps at areas where responses yield good results (if any). How we really learn languange is probably a difficult answer to derive, however, a lot of it is based on learning common phrases or responses to various situations. It would involve learning, forgetting, and memory on a continuous basis. Is it chaotic or random? eh, ... i guess i'll have to use the rest of my brain (besides the pointy end) to find out. Short answer: you are wrong, since we are at war! haha ;)

Share this comment


Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Advertisement
×

Important Information

By using GameDev.net, you agree to our community Guidelines, Terms of Use, and Privacy Policy.

GameDev.net is your game development community. Create an account for your GameDev Portfolio and participate in the largest developer community in the games industry.

Sign me up!