身体性の勉強会のメモ
Cybernetics
Computing theory of minds
アンティキティラ島の機械(Antikythera mechanism)
遠心調速機(Centrifugal Governor) (feedback system)
You have to know parameter to set the machine
Not completely auto, but Desing from outside is needed because these machine has aim human give.
アナログ計算機(Analog Computer)
our mind(?) is mechanised
Externalised our computation/mental process
one possible way of solving problem
Animal is オートマタ(Automata)
creating space of science away from gods
Integation of material and human minds
How our behavior emmerge from materials?
一元論(monism)
Cogito, ergo sum (ルネ・デカルト(René Descartes))
"I think, therefore I am"
我思う故に我あり
Many scientist ignore this
Two way of thinking mind
1. mind is illusion
2. mind is a part of nature
how?
Apply neuton physics to the animal behavior
Stability and Adaptability
Even if we interact others unusual way, they sometimes dont change their behavior.
Perturvation and Recover from it.
Interigent of sponge?
Existence requires stability
what is life about?
Cybernetics(Wiener)
Works
Information entropy
Neural Network
feedback
Positive
No final state
Negative feedback
perturvation and recorver
Stable
Final goal
Control theory
Application to war
If peoplr behaves like machine, it match to the science?
Feedback
All purposeful behavior may be considered to require negative feedback
Complex adaptive system
Warren S. McCulloch・Walter Pitts
形式ニューロン
a logical operation by NN
Why dose the burnt kitten avoid the fire?(Design for a Brain: The Origin of Adaptive Behaviour)
Only one time experience change their behavior.
Physiological limit of range
but the agent dont know the limit
Two feedback to adapt to any possible perturbation
Animal and nature has no intelligent designer.
Neuron itsself dont know of the state of whole brain and body.
No access but can control
No representation only paremeter?
Stafford Beer
designing freedom
brain of the the firm
Why people can adapt to the inverted world?
Mouse can?
Symbolic AI
Rosenblueth (1943)
Embodied AI
W. Ross Ashby(1960)
Design for a Brain: The Origin of Adaptive Behaviour
triumph of information theory
but most complex system bidirectionary affected
Computer science as empirical inquiry: symbols and search (1967)
CS is originaly from Cognitive science.
From symbolic AI to robotics
ASIMO freeze
Automata is better?
Nothing changed
Symbolic AI to Embodied AI
Naa_tsure.icon摂動を加えたときにもとある状態に戻ろうとするスポンジに知性はあるのか