Tools for Collective Human Understanding
Entertainment is the lubricant of collective adaptability.
Learning from researchers working to understand living systems is always a pleasure. Thank you to all the Santa Fe Institute’s June 2023 Collective Intelligence Symposium participants for accepting the challenge:
What is the nature of intelligence in social insect societies, adaptive matter, groups of cells like brains, sports teams, and AI, and how does it arise in these seemingly different kinds of collectives?
The following notes integrate the knowledge shared to study one of the most complex systems we know: cities. Beginning with the acknowledgment that we live inside the system we are studying–– and it does not appear possible to build a simulation of intelligence that embodies more evolutionary knowledge than life itself.
Humans build cities on millions of years of evolutionary experience and gravitational rhythms. This equates to roughly 60-400 radical climate change solar cycle events on Earth. Such events become encoded in the earth's biology. Most importantly, we live in cities and need them to support our survivability, so even without hard proof, we need to do our best to build better, more collectively intelligent cities.
1.) fundamental phenomena
How do we know collective intelligence measures are fundamental rather than nominal?
A macro-scale phenomenon is intelligent when it appears similar from different micro-scale viewpoints: when systems achieve approximate consensus or, biologically when a system achieves metabolism, and civilizationally, when places continue supporting life across centuries.
Intelligent phenomena are visible in cities through coordinated actions, combining large-scale and small-scale input when implementing infrastructure projects and their operations. Urban infrastructure includes the transportation of people and goods, communication networks, resource distribution mechanisms, symbolic constructions, and cultural exhibitions. Although many viewpoints occur, universal value is understood when operating infrastructure runs smoothly (such as infrastructure producing sanitation and clean water). It does not matter if it is a democracy or a dictatorship (gratuitous provocation to people thinking like Westerners), democracy being one method, aristocracies another, to provide system-level feedback to group-level decision-making. As historians have documented, cities collapse when feedback mechanisms become obstructed by natural or human calamities.
Before 1800, humanity had never built a city with more than 2 million people; today these are ubiquitous. As metropolises scaled up to ever larger populations (37 million in Tokyo), it was necessary to constantly update feedback mechanisms to make decisions that accommodated evermore new and different viewpoints. With such growth came evermore robust infrastructure systems to survive (pandemics) and thrive (culture).
An intelligence measure of cities would document how well society's infrastructure operations are coordinated with the needs for moving people, resources, and ideas. We hypothesize that well-coordinated transportation and communication infrastructure will produce more intelligent resource-sharing mechanisms. However, it is the human sense of robust symbolic constructions and cultural exhibitions, or collective intentionality, that measures the collective intelligence of cities.
2.) two types of measurements
How might measures be rigorously derived from microscopic, individual-level data, or by using first principle approaches?
Likely measures will be revealed as first principles of organization, and essential microscopic feedback will be measured through activity data.
In cities, measures of microscopic feedback from local users within each boundary condition of transportation and communication infrastructure can be specified in the frequency of use data relative to resource consumption. Back in the 20th century, many proposed ‘free’ mass transit services. First principles of organization can be used to structure nested levels of human activity. Animate’s Kid Cities urban development framework has seven nested levels with associated communication, transportation, and other protocols. At each level, there exist many kinds of networked organizations:
Managing dynamic metropolises today requires rigorous assessment of feedback from the constituent parts (micro-level) at each of the seven levels of description. Each of these seven levels is its own ‘complex adaptive system’ exhibiting some amount of collective intelligence. As a whole, a metropolis is a complex adaptive system level organism, coordinating many sub-components and infrastructure at each level of description. The complex adaptive system diagram by Murray Gell-Mann is an excellent place to start
The level of description is the fundamental principle of collective intelligent systems because, without it, studies become distracted by extraneous information overload. Similar to biological systems that use various permeable membranes to exclude information and enact clear boundaries of metabolism. Besides the attractiveness of patterns to the human eye, “birds, bees, flies, and fish” are valuable study systems because the boundaries separating agents from systems is obvious. In 21st-century human metropolises, these boundaries are more difficult to disentangle and empirically define.
3.) possible measurements in cities
What are the most promising measures of intelligence in brains, groups, and AI?
Metabolism, environmental work performed, understanding.
In cities the most promising measures of intelligence are minimizing petty crime, which increases human cooperation at the micro-scale, maximizing patents that perform work on the macro-scale environment, and developing a steady metabolism experienced through cultural civility at the mesoscale.
Justice is a steady-state ideal, and all governance forms attempt a specific reading of justice, even when sometimes apparently unjust from other viewpoints. This brings us back to the notion of collective intelligence appearing similar from widely different viewpoints. Using these measures of collective intelligence in cities, we see the collective consistency of sensors, spread across diverse agents within a level of description, approximates fundamental laws or parameters of behavior. Different positional viewpoints reveal diversity even with the similarity of sensor-revealed information.
Urban life in the 21st century is struggling to find common understandings caused by the unprecedented growth in population and cities that have scaled up to mega-metropolises. Humans are creating new technologies delivering “work performed on the environment,” yet many steady-state structures such as religions, social practices, and civility are losing relevance in day-to-day Western lives. One measure that would increase understanding at the mesoscale would be a scientific analysis of religions and social practices to ascertain the standard features across such practices that increase group-level survivability. Such practices have evolved over thousands of years and retain historic group-level intelligence.
4.) intelligence “scaled-up”
How does intelligence scale with other collective properties?
In nested structures.
In cities scaling of intelligence corresponds to Animate’s Kid Cities seven-level urban development framework and is evidenced in education:
What’s more, these nested structures correspond to humans' natural development levels as they age/grow. Within each level of description, say, Colleges at the mesoscale, multiple internal micro-scale organizations Grades 7-12 in this case, accommodate the diversity of thinking and methods suitable to culture, demographic, and geographic variances. These variables represent the bottom-up or top-down influences on the system depending upon the level of description. The metabolism at the level of description is the steady state consensus within an institution, our example of education in Colleges at the mesoscale, which is also influenced at the macro-scale by Universities that necessitate specific credentials for advancement.
These nested and proliferate structures are the source of ‘diversity’ at the micro-scale of agents/entities. Viewing each mesoscale level of description, evolutionary “selection cannot set in until there are entities to select.” Intelligence is measured as the coherent metabolism within the meso-scaled system. Operating in a metabolic framework, the unique view from many agents using consistent kinds and capacity of sensors, a diversity of viewpoints is created from which evolution selects, possibly refines, or minimally ‘hones’ the metabolism to environmental perturbations.
In cities these intelligence structures also correspond to physical spaces and centers of activity within boundary conditions. Poly-centric levels of activity are nested together into mega-metropolises, scale by scale, step-by-step, into multi-scaled networks of places we experience when we travel.
Today’s scaled-up metropolises need nested and scaled governance structures reflecting feedback inside 21st-century mega-metropolises. The proof of intelligence in governance is revealed in accurate feedback mechanisms and metabolism maintenance at each level‘s need for collective decision-making.
5.) general intelligence
What is general intelligence?
Intelligence is the appearance of intentionality at one scale, attributed to smaller-scale entities. It is a shared understanding between agents, species, or worlds. Metabolism is a general principle.
In cities a General Intelligence is experienced through culture. Civility is the prerequisite of freedom: one can only be free if one respects another’s right to freedom., Culture is how people treat each other in an environment. We intuitively know that people treat each other differently in small towns (or Districts) than in large metropolis-wide events. The wider environment, including climate, also has a tremendous impact on the emergence of how people treat each other. Climate forces humans to adapt, creating distinctly different cultures exhibited at the equator versus the arctic circle.
The perception of this movement humans refer to as intentionality. The metabolic framework supports unique positional views from many agents. Since all agents use consistent kinds and capacities of sensors, the collective processing refines, or minimally ‘hones’ the direction of movement or the appearance of intention.
6.) group performance principles
What principles underlie the group performance when [as] the environment is changing?
Micro-scale information processing, mesoscale coordinating activities, macro-scale environmental sensing, and resource distribution.
In cities adaptation happens at multiple time scales depending upon the influence of the larger environment, relative to the level of description one is studying. Nonetheless, a single organizational component (or component within a larger organization) operates at three levels simultaneously: macro, meso, and micro. An organization at the macro-scale completes resource distribution and wider context sensing. For example, the Chicago Black Hawks hockey team manages its resources and integrates with the National Hockey League. When operating at the mesoscale, the same organization maintains a metabolism that coordinates the activities of its constituent organizations. The Black Hawks administration and staff have done this routinely for almost 100 years. And finally, an organization's constituents, when acting at its micro-scale level, process information regarding energy and environment: the Chicago Black Hawks players and coaches.
Coordination done well, and winning NHL Stanley Cups, is difficult.
Judgment describes a decision made for the collective, with strong consideration given to the long-term and broad viewpoints of said collective. Judgment is the enactment of the intended group performance. Only repetitive assessments from different viewpoints and across time can evaluate the performance of a judgment.
7.) adaptation capacities
What measures predict the capacity of a collective to adapt?
Intermezzo:
Entertainment is the lubricant of collective adaptability.
In cities, humans use entertainment to break symmetry, release energy and exchange alternative understandings of patterns. Entertainment assumes coordination: an agreement of rules to play a game, or in the case of comedy, a common situation to which a comedian reveals an ironic, transgressed, or mismatched reference. Music entertains us through resonance with fundamental frequencies in the specific world people experience., Entertainment is a low-cost means to test adaptation (like text-to-image generators or ‘oat milk horchata cold brew coffee’).
By contrast, collectives define weeds as purposeless novelty because they can be harmful. Metabolism keeps weeds out, like cancer cells and parasites. The term ‘weed’ is relative to a level of description. In another metabolism, dandelions become produce for healthy diets. Adaptation by the collective includes a judgment about survivability.
Humans do select, and we do contribute to evolution through our collective decision-making. Governance and decision-making in collectives are pretty clear about selecting the fittest: China has had a national exam for thousands of years. Collective metabolism in the West changed after the Enlightenment, the Romantics, and culminating with purely free individuals.
Possibly anti-collective, such individualism curiously coincided with colonialism. Many Western cities today are forced to adapt to vastly larger scales without stable metabolisms. Creating more robust urban metabolisms requires delving into multi-scaled urban life, finding appropriate delineations at each level of description, and establishing semi-permeable boundary conditions supporting new cultural norms.
Testing a definition:
weed | wēd |
noun
a wild plant growing where it is not wanted and in competition with cultivated plants: keep the seedlings clear of weeds.
British informal, derogatory a contemptibly feeble person: he thought party games were for weeds and wets.
informal a leggy, loosely built horse: my tiny bay weed could jump like a stag.
in AI & ALife an agent with the same sensors as others, which inconsistently processes information relative to the local metabolism: that dude (weed) is not contributing to the local pattern.
In cities, four measures predict the capacity of a collective to adapt:
Nested structure of multi-scaled places, where each place has a level of description and governance/decision-making mechanism
Permeable boundary at each physical place
Visible center where repetitive interactions occur at each place
Cultural context facilitating interaction, entertainment, and cooperation inside each place
In cities, multi-scaled places with clear activity centers and permeable boundaries enable people and organizations to adapt. Multi-scaled operations of a metropolis, supported by entertainment, are conducted in a steady atmosphere within each place. In such arrangements, new information flows to specific scaled places. Repetitive interactions by people and organizations facilitate the steady cultural development of cooperation. Intelligent adaptation happens when new information flows and is processed at the appropriate/effective level of description.
A well-defined level of description supports governance and resource distribution mechanisms to be more effective.
Healthy metabolisms enable adaptation without sacrificing the host.
Judgments are made continuously by artists. Every brush stroke is a judgment made through muscle-memory and philosophical meanderings, which conspire into an “impression” that resonates across space, time, and cultures. Paintings, as a form of entertainment, encourage humans to see things differently, contributing to a more accurate understanding of our universe.
A note: Georges Seurat’s refined style can be replicated with an algorithm. In comparison, Van Gogh’s sense of composition, subject matter, and technique combine into something beyond artificial intelligence.
And a final note about humans.
What amazes us (humans) is that as far as we know, we are at the top of the complexity pile. We know we are composed of atoms, cells, organs, bones, blood vessels, and a brain. Moreover, we know this is organized by metabolism at 98ºF. And we know that the foods we eat affect our thinking. We know that we care about our family and friends. We know that we interact with our colleagues and participate in multi-scaled, overlapping organizations like the United States, Chicago Black Hawks, Petanque Clubs, political parties, community clubs, professional associations, and global NGOs that care for the planet.
No other animals traverse the scales of information processing that humans do. Nevertheless, perhaps exposed to the fear of AI, ‘there will be more.’ As evolution continues, there will be more complex life forms than us humans. The next evolutionary speciation into a more complex life form will need more sensors than we can absorb and more robust information processing systems to accommodate more types of sensory data.
The ability of humans to process information at so many scales also obscures the value of well-organized levels of organization and governance. While traversing many scales of information processing can be done by single individuals, often called heroic, such processes short-circuit general intelligence by overlooking potential contributions from outsiders. Portland, Oregon's Commission form of governance, unlike any other large American metropolis, operates precisely without any governance levels for smaller Districts, Neighborhoods, and Villages. As such, it limits the formation of local viewpoints, contributions, and multi-scaled metabolism. A lack of influence may have inadvertently nurtured an anti-establishment, transgressive motivation in Portland’s environment.
Adapting the governance structures to advanced sensors and feedback, as well as increased scale forcing a phase change (i.e., French revolution) is an indicator of a higher level of collective intelligence. Human’s ability to traverse the scales of information processing becomes critical during a phase change, rapidly changing sensibilities of what is possible. New metabolisms are formed by consistent kinds of sensors finding multi-scaled local boundary conditions with diverse positions and viewpoints.
What a ride…
