The Society of Mind by Marvin Minsky Motivation

The Society of Mind by Marvin Minsky

Motivation What is human mind and how does it work? How do we recognize objects and scenes? How do we use words and languages. How do we achieve goals? How do we learn ? How does “common sense” work?

Motivation(contd. ) An individual ant is not very bright, but ants in a colony, operating as a collective, do remarkable things. "A single neuron in the human brain can respond only to what the neurons connected to it are doing, but all of them together can be Albert Einstein. " By Deborah M. Gordon (Stanford University)

Analogy : Ants' society Some social systems in Nature can present an intelligent collective behavior although they are composed by simple individuals. The intelligent solutions to problems naturally “emerge” from the self-organization and communication of these individuals.

Analogy : Ants society(cont. ) Individual ants are simple insects with limited memory and capable of performing simple actions. However, an ant colony expresses a complex collective behavior providing intelligent solutions to problems such as: carrying large items forming bridges finding the shortest routes from the nest to a food source, prioritizing food sources based on their distance and ease of access.

Analogy : Ants society(cont. )

Content Introduction K-lines, Nomes, Nemes K-lines example, Large Agencies, Problem Solving Communication and Growth of Mental Societies Recent developments and conclusion

Agents Any part or process of Brain that by itself is simple enough to understand Agents constitute the building blocks of the mind In society of mind, mental activity reduces to turning individual agents on and off Unintelligent

Agency Society of simple agents Can perform functions more complex than any single agent could Complicated behavior is the result of the interaction between groups of agents

K-lines Turns on a particular set of agents Activating a K-line can cause a cacade of effects in the mind Reactivates the previous mental state based on the similarities between current situation and the situation previously encountered Causes to enter particular remembered configuration of agents

Classes of K-lines Nemes are concerned with the representing aspects of the world Nomes are concerned with controlling how the representations are manipulated.

Nemes produced by learning from experience Polynemes invoke partial states within multiple agencies polynemes supports that meaning can be better expressed across multiple representations

Micronemes Refers aspects of a situation that are difficult to attach to any particular thing Feelings, emotions etc

Nomes Controls how representations are manipulated Isonomes signal to different agencies to perform the same uniform type of coginitive operation Pronomes control the use of short-time memory Pronomes are often associated with a specific role in a large situation or event

Paranomes Set of pronomes linked to each other Changes made by one pronome produce corresponding changes by other pronomes to related representations

Example of K-line attached to many agents.

Example of K-line formation

K-line formed due to the event

How to combine agents to form larger agencies? Frames are a form of knowledge representation Concerned with representing a thing and all the other things or properties that relate to it in certain particular ways. Has slots. Built from pronomes which control use of slots. Frame-arrays Collection of frames which share slots.

How to combine agents to form larger agencies? Each frame describes the thing from some particular perspective or point of view. Transframes Represent events and all of the entities that were involved with or related to the event. Picture-frames that represent the spatial layout of objects within scenes.

How can agents solve problems? Difference Engines Simple Machine. Operate on difference between the current state and goal state. invoke suitable k-lines which reduce the difference. Censors and Suppressors Additional Knowledge about common pit falls and bugs in the methods. Supress the actions which lead to bugs.

How can agents solve problems? (contd. . . ) A-Brain and B-Brain Some type of pitfalls not particular to any method looping B-Brain think about the A-Brain.

How do agents communicate with each other? K-Lines invoke other agents in the mind. Connection-lines agents Internal connected to each other through bus. language Similar to how people communicate with each other in their natural language. Paranomes Common Method

Growth of Mental Societies Protospecialists Highly evolved agencies that produce behaviors provide initial solutions to problems such as locomotion, obtaining food. Predestined Learning Abilities which are shared among all the people. eg. walking

Growth of Mental Societies Types of Learning Accumulating Remember each example or experience as separate case. Uniframing Finding general discription which subsumes multiple examples. Transframing Anology or some other form of bridge between two represenatations. Reformulation New ways to describe the existing knowledge.

Growth of Mental Societies Learning from attachment figures How to learn goals in the first place. Interaction with attachment figures. Learning from Mental Managers mental growth are based not simply on acquiring new skills acquiring new administrative ways to use what one already knows Development Multiple Stages. Train each other.

Recent Developments Case-based This reasoning field studies implementation details of Klines. The methods that have been developed by this community are the most similar in spirit to Minsky's ideas about K-lines

• Multiagent systems This Field tries to answer the kinds of questions about how one might build a Society of Mind. Researchers have proposed many ideas about how agents should communicate, how they might coordinate their different goals, how they might work together to plan solutions to problems, and so forth. There are now a wide variety of architectural ideas about how to build multiagent systems

Consequences of theory Godel's Incompleteness Theorem. Technological Singularity. Artificial Creativity.

Are People Machines? Human activities can be broken down into small, simple actions. It may be that humans are only highly complicated machines with billions of parts.

References [1] Push Singh: Examining the Society of Mind. Computers and Artificial Intelligence 22(6): (2003) [2] Marvin Minskey: Society of Mind. Book [3] Julie Minskey: Society of Mind. Article. www. oss. net/dynamaster/file_archive/0403 24/e 4 bbee 1 dc 1385 f 85730 df 870 d 1608682/ WER-INFO-64. pdf [4] Artificial Consciousness. Article. www. alextreme. org/docs/faai/Artificial_Con sciousness-24042004. html