Creative Problem-Solving Swarm

Idea Swarm

Inertia w (momentum) 0.60

Personal pull c₁ (own best idea) 1.4

Social pull c₂ (group consensus) 1.4

Search State

Phase Diverging

ConvergedDiverse

Iterations

Best idea score

Peaks in problem

Diversity

How it works

Each glowing sphere is a candidate "idea" living on a random fitness landscape (the terrain) — height = how good that idea is. A swarm of ideas searches the landscape using particle-swarm optimisation (PSO), a computational model often used to reason about how divergent and convergent thinking combine during brainstorming:

v ← w·v + c₁·r₁·(pBest − x) + c₂·r₂·(gBest − x)
x ← x + v·Δt

w (inertia) — how much an idea keeps drifting in its current direction. High w keeps the swarm diverging — exploring wildly different regions, like free brainstorming.
c₁ (personal pull) — how strongly an idea is drawn back toward the best variant it personally has found so far.
c₂ (social pull) — how strongly every idea is drawn toward the best idea anyone in the group has found. High c₂ drives rapid convergence onto a shared solution.
r₁, r₂ — fresh random numbers each step, so no two idea-trajectories are identical — the same stochastic spark behind real creative variation.

The diversity meter tracks the swarm's average distance from its own centroid: high diversity is the divergent-thinking phase (many rough ideas explored in parallel); as c₂ pulls everyone toward the best-known peak, diversity collapses and the swarm enters the convergent phase, refining one strong solution. New Problem reshuffles the landscape's peaks — a fresh problem to solve — while Reset Swarm keeps the same landscape and starts a new search over it.

Idea (particle) Best idea found

⚠ Couldn't load the 3D engineThree.js failed to load from the CDN. Check your connection and reload.