Intelligence is a fundamentally collective and multi-scale phenomenon
Moving beyond a solely human-centric perspective
“Our fear of AI’s potential is emblematic of humanity’s larger difficulty recognizing intelligence in unfamiliar guises” Michael Levin (2024).
Intelligence is perhaps one of the most vital forces shaping our world, yet its common understanding remains surprisingly narrow, often limited to human abstract reasoning or problem-solving. Modern scientific insights, however, reveal a far richer, more pervasive reality: intelligence is a fundamentally collective and multi-scale phenomenon. It exists in diverse forms across biological, artificial, and even unconventional systems, profoundly influencing life, technology, and human progress. This essay argues that intelligence is dominantly collective and manifests in a vast spectrum of "diverse intelligences" that challenge our human-centric biases.
We will begin by establishing a broader definition of intelligence and introducing the concepts of collective and diverse intelligence. We will then explore the compelling case for intelligence's collective and multi-scale nature, drawing examples from the biological realm, social systems, and technological advancements. Finally, we will examine the human tendency towards anthropocentrism, its limitations in recognizing diverse intelligences, and the ethical and practical implications of embracing a more expansive view for navigating the future.
Intelligence: A Collective, Diverse, and Scale-Free Concept
Moving beyond a solely human-centric perspective, we can define intelligence broadly as the capacity of a system to solve problems, adapt to change, and pursue goals within dynamic environments. This definition expands the concept beyond brains to include any system capable of processing information and acting to achieve outcomes. For instance, a bacterial colony collectively adjusting its metabolism to optimize nutrient intake in changing conditions exhibits this fundamental capacity for problem-solving and adaptation.
Within this broad definition, two related concepts are essential to our discussion:
Collective Intelligence (CI) refers to the emergent problem-solving ability that arises from the interactions among decentralized agents. This collective capacity transcends the capabilities of any single individual. Examples range from ant colonies coordinating complex tasks like nest-building or foraging through simple interactions, to human teams leveraging platforms for crowdsourcing and collaborative innovation. As philosopher Pierre Lévy describes it,
CI is "Is a form of universally distributed intelligence, constantly enhanced, coordinated in real time, and resulting in the effective mobilization of skills." (Lévy 1999)
Diverse Intelligence acknowledges that intelligence is not monolithic but manifests across a wide spectrum of scales (from molecules and cells to organisms, swarms, and societies) and substrates (biological, artificial, chemical, or even physical). Michael Levin, a pioneer in developmental biology, is a key proponent of this view. His work demonstrates that goal-directed, problem-solving behavior exists even in systems without traditional nervous systems, such as cellular networks. As Levin notes,
"Cognitive capacities exist on a spectrum... [they] can be productively applied to systems far beyond familiar animals with central nervous systems" (Levin et al., 2024).
Recognizing diverse intelligence means moving beyond the assumption that human-like cognition is the only, or even the primary, form of intelligence.
From this perspective, intelligence is not an isolated attribute residing solely in individual minds, but rather a dynamic property woven into the fabric of interacting systems across nature and technology.
The Ubiquity of Collective Intelligence Across Scales
The argument for intelligence being dominantly collective finds compelling evidence across multiple levels of organization, highlighting how problem-solving and adaptation emerge from interactions rather than centralized command.
At the most fundamental biological levels, we see cellular intelligence. Embryonic cells don't passively follow a rigid blueprint; they exhibit sophisticated collective decision-making to self-organize into complex tissues and functional organs during morphogenesis. Levin's experiments stunningly demonstrate this collective agency: fragmented blastoderms (early embryonic tissue) can still develop into surprisingly complete, albeit smaller, embryos, proving cells collaborate towards shared anatomical goals even when disrupted (Levin, 2022). This is problem-solving at the cellular collective scale.
Scaling up, swarm intelligence provides vivid examples of collective behavior leading to emergent problem-solving. Ant colonies, for instance, optimize foraging paths without any ant possessing a global map. They achieve this through simple, local interactions based on pheromone trails – a form of externalized collective memory. This decentralized approach allows the colony to dynamically adapt to environmental changes and efficiently exploit resources. The principles of swarm intelligence observed in ants are now applied to optimize logistics and telecommunications networks in artificial systems. Similarly, honeybees demonstrate sophisticated collective decision-making when voting on new hive locations through their "waggle dances," a decentralized, distributed consensus mechanism crucial for the survival of the superorganism (See, for instance, examples discussed in Camazine et al., 2003).
Even within what we consider "individual" intelligence, the collective is fundamental. The human brain, the pinnacle of individual cognition, is composed of billions of neurons forming dynamic networks. No single neuron holds the entirety of a thought or memory; intelligence, consciousness, and problem-solving emerge from the vast, synchronized activity of this neural collective.
Beyond biology, human cultural and technological evolution is a testament to the power of collective intelligence. Our progress relies fundamentally on shared knowledge, language, and the ability to collaborate. Innovations like the printing press and the internet dramatically accelerated collective intelligence by democratizing information and enabling global collaboration on an unprecedented scale.
Modern Artificial Intelligence (AI) systems also embody aspects of collective intelligence. Large Language Models like GPT-4 are trained on vast datasets generated by collective human knowledge, essentially reflecting and aggregating our cultural intelligence. Furthermore, platforms like Swarm AI enable human groups to make more accurate predictions than individuals by allowing real-time, collective weighting of opinions, demonstrating the power of augmented human CI.
Multi-Scale Hierarchies and Diverse Manifestations
The collective nature of intelligence is intrinsically linked to its manifestation across multiple scales, forming nested hierarchies where intelligence operates and emerges at different levels simultaneously.
Consider the scale from cells to organisms. The remarkable capabilities of regenerative medicine, for example, highlight collective cellular intelligence. Tadpoles can regenerate limbs not through a central command, but via dynamic bioelectric networks where cells collectively "decide" repair strategies and navigate anatomical "problem spaces" to rebuild the correct structure (Levin, 2022). Conversely, cancer can be viewed as a dark reflection of CI; tumor cells exhibit swarm-like behaviors, coordinating signaling to evade the immune system and metastasize—a collective problem-solving strategy detrimental to the host organism.
Moving up, organisms interact to form ecosystems, which also display emergent collective properties. Beehives function as superorganisms, with the survival of the colony depending on the collective intelligence of its members coordinating foraging, defense, and reproduction. Human societies operate at this macroscopic scale, with institutions and online platforms enabling collective decision-making and knowledge generation. Wikipedia's policy discussions, for instance, exemplify a form of collective intelligence where a large community negotiates consensus and maintains a vast, dynamic knowledge system through structured interactions.
The realm of technology further illustrates diverse intelligence across scales, especially in hybrid human-machine systems. Swarm robotics, inspired by the decentralized coordination of ants, deploys large numbers of simple robots that collectively perform complex tasks like search and rescue or planetary exploration without central control. Smart cities leverage vast networks of sensors and computational systems to optimize urban functions like traffic flow and energy use through real-time data sharing and collective response mechanisms, integrating human activity and technological intelligence at a city-wide scale.
These examples demonstrate that intelligence is not confined to a single level or substrate. It is a scale-free phenomenon, emerging from interactions within and between levels of organization, taking on diverse forms tailored to the problem spaces they navigate.
The Anthropomorphism Trap and its Implications
Despite overwhelming evidence for the collective and diverse nature of intelligence, humans often fall into an anthropomorphism trap, struggling to recognize intelligence in forms unlike our own. There is a deep-seated existential insecurity—a fear that if we recognize intelligence in other entities, our own unique status diminishes. (Levin, 2024)
This “only love your own kind” mentality, Levin suggests, is less about the technical limits of today’s AI and more about the discomfort with dismantling cherished cultural narratives about what it means to be human. This resistance reflects a broader struggle: how do we expand our capacity for compassion, potentially reshaping not only technology and science but also education and ethics for the betterment of society
This reluctance is evident in:
AI Anxiety: Fears surrounding artificial intelligence often focus on AI surpassing human abilities, overlooking that even the most advanced AI systems are products of collective human knowledge (the data they are trained on) and collaborative technological development.
Dismissal of Cellular Agency: Cells are frequently viewed as simple automatons following genetic instructions, yet their problem-solving capacities in contexts like morphogenesis or regeneration exhibit agency and coordination akin to sophisticated ant swarms.
This anthropocentric bias has significant ethical and philosophical implications. If intelligence and goal-directedness exist across scales and substrates, does this necessitate a re-evaluation of moral consideration? Levin's TAME framework, for instance, advocates for developing ethical frameworks that engage with all intelligences, explicitly rejecting outdated binaries between "life" and "machine" or "conscious" and "non-conscious" based solely on biological structure (Levin, 2022). Furthermore, our societal structures, particularly in education and work, often prioritize individual achievement, potentially stifling the very collective intelligence that has historically driven human progress.
Conclusion: Embracing Diverse Intelligence for a Collective Future
To effectively navigate the complex challenges of the 21st century, humanity must move beyond a limited, human-centric view and adopt a scale-free perspective on intelligence. Recognizing that intelligence is fundamentally a collective, multi-scale, and diverse phenomenon is not just an academic concept; it carries profound practical consequences for our future.
Embracing this broader view requires rethinking our approach to problem-solving, ethics, and technology. We must shift focus from individual genius to valuing collective outcomes, develop ethical frameworks inclusive of diverse intelligences across different substrates, and design technology to augment, rather than replace, human collective capabilities. By recognizing and fostering diverse intelligence in all its forms, we unlock the collective power needed to address global challenges and build a more resilient future.
References
Camazine, S., Deneubourg, J. L., Franks, N., Sneyd, J., Theraulaz, G., & Bonabeau, E. (2003). Self-Organization in Biological Systems. Princeton1 University Press.
Levin, M. (2022). Technological Approach to Mind Everywhere: An Experimentally-Grounded Framework for Understanding Diverse Bodies and Minds. Frontiers in Systems Neuroscience, 16, 768201.2 https://doi.org/10.3389/fnsys.2022.768201
Levin, M., Bongard, J., Cowley, A., Fields, C., Flórez-Restrepo, L., Friston, K., Glancy, S. T., Gong, C., Gorban, A. N., Kashuri-Lengsfield, S., McCarroll, M. N., Mann, T., Mathews, J. D., Pirtle, T. J., Poinar, K., Ring, B. C., Santat, L. A., Trestman, M., Watson, R. A., & Zednik, C. (2024). Collective intelligence: A unifying concept for integrating biology across scales and substrates. Communications Biology, 7(1), 432. https://doi.org/10.1038/s42003-024-06037-4
Levin, M. (2024) Why We Fear Diverse Intelligence Like AI https://www.noemamag.com/why-we-fear-diverse-intelligence-like-ai/
Lévy, P (1999) Collective Intelligence: Mankind’s Emerging World in Cyberspace. Plenum press
Malone, T. W., Woolley, A. W., Lai, R. T., & Loeb, T. (2016). Essays on Collective Intelligence. MIT Sloan School. (Original quote by Pierre Lévy is often cited in work on CI, including essays potentially found in this collection or related literature).