Signal
Signal#
In theory, you need about five good inputs to make a decision. In practice, you’re swimming in hundreds. Headlines, metrics, opinions, forecasts, hot takes, polls, rankings, trends — all arriving simultaneously, all demanding attention, most of them irrelevant.
This piece tries to make that experience visible.
View fullscreen — best watched through a full cycle
The Idea#
The animation moves through four stages.
Flood. Ninety particles stream in from the edges, each carrying a word: trending, clickbait, speculation, outrage, forecast, algorithm. They arrive staggered, accumulating into a dense cloud. Text flashes on and off individual particles — readable for a moment, then replaced by abstract dots. Five signal particles are hidden in the flood, indistinguishable from noise.
Saturation. The canvas is full. Words flash past too quickly to evaluate. Everything looks the same: muted blue-grey text at the same size and brightness. This is the state most of us live in — surrounded by information, unable to sort it.
Filtering. The signal begins to separate. Five particles slow down, drift toward center, grow, and shift from blue-grey to warm gold. Their words become permanently readable: evidence, context, values, experience, intuition. Meanwhile, the noise dims and pushes outward.
But filtering isn’t clean. Three noise particles briefly flash golden — poll, or ranking, or narrative — mimicking signal before being correctly identified and dimming back to noise. These misidentification moments are the most important part of the piece. In real decision-making, the danger isn’t the noise you ignore. It’s the noise you mistake for signal.
Clarity. The five signal words settle into a pentagon at center, connected by subtle golden lines. They form a framework — not a conclusion, but the inputs you actually need: what do you know (evidence), what’s the situation (context), what matters to you (values), what have you learned (experience), and what does your gut say (intuition). The noise is still visible at the edges. It hasn’t gone away. It’s just no longer in charge.
The Five Inputs#
I chose these five because they span the range of legitimate decision-making inputs:
- Evidence — external, verifiable, data-based
- Context — situational, specific, often overlooked
- Values — internal, stable, deeply personal
- Experience — accumulated, pattern-based, earned
- Intuition — pre-verbal, integrative, often right for reasons you can’t articulate
None of them is “trending.” None of them arrives as a notification.
Building It#
The particle system uses 95 particles positioned via Fibonacci spiral distribution with jitter. Each particle transitions between four positions (source → noise → signal/edge) using phase-based interpolation, making the motion deterministic and loopable.
The hybrid rendering switches each particle between readable text and abstract dots based on a sinusoidal flash function. During flood and saturation, text appears on roughly two-thirds of particles at any moment, creating a flickering data-stream effect. During filtering, signal particles transition permanently to text while noise transitions permanently to dots.
The misidentification targets are randomly selected per cycle — three noise particles whose index matches a random set. During the filter phase, they briefly adopt signal-like properties (golden hue, larger font, higher alpha) before reverting.