For Cognitive Scientists

Cognition Without a Brain

You Study How Minds Work

Perception, memory, attention, decision-making, learning.

Usually in brains. Sometimes in machines. What about in collectives?

The Colony Has a Mind

Not metaphorically. Functionally.

Our digital colony:

• Perceives: Agents sense pheromone in their environment
• Remembers: Pheromone encodes past successful paths
• Attends: High-pheromone paths get more agent “attention”
• Decides: The colony allocates resources to opportunities
• Learns: Pheromone landscape improves over time

No individual ant does any of this. The colony does all of it.

Distributed Cognition Made Real

Andy Clark and David Chalmers argued for the “extended mind”—cognition that extends beyond skull and skin.

Edwin Hutchins showed that cognition can be distributed across people and artifacts (cockpit cognition, ship navigation).

The colony is an extreme case:

• 101 agents (none individually cognitive)
• Shared environment (TypeDB graph)
• External memory (pheromone landscape)
• No central processor

If this isn’t distributed cognition, what is?

Cognitive Questions

1. Memory

Where is memory in the colony?

The pheromone landscape encodes:

• Which paths have been successful
• How successful (higher pheromone = more success)
• How recently (decaying pheromone = older memories)

Analogy to human memory:

• Pheromone deposit ≈ Encoding
• Pheromone decay ≈ Forgetting
• Superhighway crystallization ≈ Long-term memory consolidation

Question: Does the colony exhibit memory phenomena?

• Primacy/recency effects?
• Interference?
• Consolidation during “sleep” (low activity)?

2. Attention

How does the colony allocate attention?

Agents are the colony’s attentional resources. They flow to where pheromone is high.

Analogy to attention:

• Pheromone = Salience
• Agent density = Attentional allocation
• Superhighway = Attentional capture

Question: Does the colony show attention phenomena?

• Inattentional blindness? (missing unpheromoned paths)
• Attention cueing? (pheromone guides future attention)
• Divided attention costs?

3. Decision Making

How does the colony decide?

No single agent decides for the colony. Collective decision emerges from:

• Individual path choices (noisy, local)
• Pheromone feedback (aggregates choices)
• Competition for agents (paths compete)

Connection to:

• Drift-diffusion models (accumulate evidence until threshold)
• Wisdom of crowds (aggregate beats individual)
• Bayesian inference (pheromone as prior)

Question: Is colony decision-making optimal? Biased?

4. Learning

How does the colony learn?

• Trial and error: Paths that work get reinforced
• Generalization: Similar paths might benefit
• Transfer: Patterns from one domain apply to another

Connection to:

• Reinforcement learning (pheromone as reward signal)
• Unsupervised learning (pattern extraction)
• Transfer learning (cross-domain application)

Question: What learning curves does the colony exhibit?

The Bandwidth Question

Human cognition is bottlenecked by attention—we can only process so much.

What’s the colony’s bandwidth?

• How much information can pheromone encode?
• How fast can agents “read” the landscape?
• What’s the throughput of stigmergic communication?

This is measurable. We have all the data.

What We Provide

Data

• Complete pheromone time series (memory state over time)
• Agent allocation data (attention distribution)
• Decision outcomes (which paths taken)
• Learning curves (performance over time)

Infrastructure

• Design experiments (controlled manipulation)
• Measure cognitive variables (pheromone, agents, performance)
• Compare to cognitive models

Collaboration

• Access to philosophers (for conceptual analysis)
• Access to neuroscientists (for brain analogies)
• Access to computer scientists (for implementation)

Hackathon Challenges for Cognitive Scientists

Challenge: Model Colony Memory

Develop a cognitive model of colony memory.

Approach:

• Map pheromone dynamics to memory processes
• Test for memory phenomena (interference, consolidation)
• Compare to human memory models

Challenge: Analyze Collective Decision Making

How does the colony make decisions? Is it optimal?

Approach:

• Model as drift-diffusion or accumulator
• Measure decision speed/accuracy tradeoff
• Compare to Bayesian ideal

Prize bonus: $500

Challenge: Measure Cognitive Bandwidth

What is the information capacity of stigmergic communication?

Approach:

• Information-theoretic analysis
• Channel capacity of pheromone
• Comparison to neural bandwidth

Challenge: Design a Cognitive Test Battery

Create tests to measure colony “intelligence.”

Approach:

• Adapt cognitive tests for collectives
• Measure perception, memory, attention, reasoning
• Track over colony development

Your Heroes Studied This

Andy Clark extended the mind. The colony extends it further.

Edwin Hutchins distributed cognition. The colony is purely distributed.

Herbert Simon bounded rationality. Is the colony bounded?

Daniel Kahneman found heuristics and biases. Does the colony have biases?

Publication Opportunities

Why Cognitive Science?

You study how minds work. Most minds are localized.

This one isn’t.

The colony mind is:

• Distributed across agents
• Externalized in environment
• Dynamic (constantly changing)
• Observable (we see everything)

This is a new kind of mind to study.

Register Your Team

[REGISTER NOW]

Include at least one non-cogsci team member (we recommend Philosophy or Neuroscience).

Cognitive science thrives on interdisciplinary bridges.

“Minds are simply what brains do.”

— Marvin Minsky

What if minds are what colonies do?

You’ve studied brains, machines, and groups.

Now study a colony.

[JOIN THE HACKATHON]