The Society of Mind
In plain wordsAI pioneer Marvin Minsky's 1986 book arguing that intelligence comes from huge numbers of tiny, simple processes working together, no one of which is smart by itself.
Marvin MinskyMarvin Minsky, prop. n.: An American AI pioneer who helped launch the field and MIT's AI laboratory, co-wrote a famous critique of early neural networks, and pictured the mind as a society of many simple parts.'s 1986 book, and the theory of mind it lays out, on which intelligence is not a single faculty but the joint product of a vast number of simple processes — Minsky's 'agents' — each too simple to understand anything by itself. Agents combine into agencies that compete, cooperate and suppress one another; memories are explained by K-lines, which reactivate the set of agents that were active when a problem was solved, and structures such as framesframe, n.: In classic AI, a ready-made template for a familiar kind of thing or event, such as a kitchen or a job interview, with blanks to fill in and sensible default values. organize expectations. Written as some 270 one-page essays, the book is a sketch rather than a formal or implemented model, and it argues against a unified self or central controller. It influenced research on multi-agent systemsmulti-agent system, n.: A setup in which several AI agents, each with its own role or point of view, work together — or compete — on a problem, passing messages and results among themselves., behavior-based robotics and cognitive architecturescognitive architecture, n.: A blueprint for a complete thinking system — how memory, attention, learning and decision-making fit together — used both to build general AI agents and to model how the human mind works., and its title is often borrowed for language-modellarge language model, n.: A computer program trained on enormous amounts of text to predict the words that come next. At vast scale, that simple skill lets it chat, answer questions, translate, summarize and write code. systems in which several agents debate or collaborate on a task.
The paper frames its team of debating LLM agents as a society of mind, though Minsky's agents were far simpler.
Origin
Developed by Marvin Minsky with Seymour Papert at MIT from the early 1970s, partly prompted by attempts to make a robot arm and camera build towers of children's blocks, and published by Simon & Schuster in 1986. Minsky extended the theory to emotions and reflective thinking in The Emotion Machine (2006).
Usage note
Minsky's agents are internal mental processes, not the autonomous software AI agentsAI agent, n.: An AI system that does not just answer but acts: given a goal, it decides on steps, uses tools such as a browser or code editor, checks the results and keeps going until the job is done. of current usage; multi-agent LLM papers use the phrase loosely.
Related
- Marvin MinskyMarvin Minsky, prop. n.: An American AI pioneer who helped launch the field and MIT's AI laboratory, co-wrote a famous critique of early neural networks, and pictured the mind as a society of many simple parts.
- multi-agent systemmulti-agent system, n.: A setup in which several AI agents, each with its own role or point of view, work together — or compete — on a problem, passing messages and results among themselves.
- frameframe, n.: In classic AI, a ready-made template for a familiar kind of thing or event, such as a kitchen or a job interview, with blanks to fill in and sensible default values.
- cognitive architecturecognitive architecture, n.: A blueprint for a complete thinking system — how memory, attention, learning and decision-making fit together — used both to build general AI agents and to model how the human mind works.
- symbolic AIsymbolic AI, n.: An approach to AI that writes knowledge down as human-readable symbols and rules, such as 'every bird has wings', and reasons by manipulating them step by step instead of learning patterns from data.
- global workspace theoryglobal workspace theory, n.: A leading scientific theory holding that we become conscious of something when the brain broadcasts it widely — as if putting it on a stage where every part of the mind can use it.
Sources
- Minsky, M. (1986). The Society of Mind. Simon & Schuster.
- Du, Y., Li, S., Torralba, A., Tenenbaum, J. B. & Mordatch, I. (2023). Improving Factuality and Reasoning in Language Models through Multiagent Debate. arXiv:2305.14325.