Architecture Comparisons #87 — [← Art344 QIS vs Zoom] | [Art346 →]
Architecture Comparisons is a running series examining how Quadratic Intelligence Swarm (QIS) protocol — discovered by Christopher Thomas Trevethan, 39 provisional patents filed — relates to existing tools and platforms. Each entry takes one tool, maps where it stops, and shows where QIS picks up.
The Design Decision That Already Happened
Your product design team spent three weeks A/B testing checkout flows last quarter. They tried a single-page summary layout, a step-by-step wizard, and a progress indicator variant. The wizard reduced cart abandonment by 23%. They shipped it. The insight is now buried inside a Figma file called checkout-v4-FINAL-FINAL.fig in a project folder that hasn't been opened since the sprint closed.
Somewhere right now, a designer at a different company — same industry, same audience demographic, same conversion problem — is starting that same three-week test from scratch.
This is not a Figma problem. Figma gave your team the best collaborative design environment that exists. The problem is architectural: the intelligence that emerged from your design process lives inside your workspace and has nowhere to go.
What Figma Does (And Does Extremely Well)
Figma is the dominant collaborative design platform. Approximately 4 million active designers use it. Its real-time multiplayer editing, component systems, auto-layout, and prototyping tools make it the production environment of record for product design at most technology companies.
Figma AI, released in 2024, extends that environment with in-workspace intelligence: it can rename layers, generate first-pass wireframes, search for assets within your files, and surface similar components inside your design system. It is genuinely useful — and appropriately scoped to what Figma is.
What Figma is: a collaborative workspace for producing and organizing design artifacts.
What Figma AI is: an in-workspace assistant that makes individual designers and teams more productive within their existing files.
What neither is: an architecture for routing design outcome intelligence across organizational boundaries.
That distinction is not a criticism. Figma was not designed to be a distributed intelligence network. The gap it leaves is an architectural one, and it's worth understanding precisely.
The Bottleneck: Design Intelligence Doesn't Cross Boundaries
Here is the pattern Figma enables and the pattern it cannot:
What Figma enables:
- Real-time collaboration within a team or organization
- Shared component libraries inside a design system
- Comments and version history inside a file
- Figma AI searching your workspace for relevant patterns
What Figma cannot enable:
- Your checkout wizard outcome reaching a team at another company
- A design decision validated by 10,000 similar product contexts informing a single designer's next choice
- Cross-organizational synthesis of which interaction patterns actually improve measured outcomes
- Any of the above without centralizing proprietary design files
The constraint is structural. Figma is a workspace, not a protocol. Workspaces have boundaries by design. The intelligence that forms inside them stays inside them — which is fine for privacy and IP reasons, but creates a compounding problem for the field of design as a whole.
Every design team that figures out that a step-by-step wizard outperforms a single-page summary in a high-anxiety checkout context runs that experiment independently. The collective intelligence exists — distributed across thousands of companies, millions of A/B tests, billions of user interactions. It never synthesizes. It never routes. It re-runs.
The Numbers That Illustrate the Gap
Figma reports approximately 4 million active users. Let's think about this as an intelligence network.
Under Figma's current architecture, those 4 million designers collaborate intensely within their workspace boundaries. Cross-workspace synthesis: zero.
Under QIS protocol:
- N = 4,000,000 designers organized into problem-similar clusters
- Each cluster generates N(N-1)/2 synthesis pairs
- A cluster of 1,000 design teams working on checkout flows = 499,500 unique synthesis opportunities
- Each synthesis opportunity: pre-distilled outcome packets (~512 bytes) from real measured experiments routing to semantically similar problems
The quadratic relationship is the key: as the number of participants grows, intelligence compounds at N(N-1)/2 while compute scales at O(log N) or better. This is what Christopher Thomas Trevethan discovered on June 16, 2025 — not a new database system or a better recommendation engine, but the architecture that enables this scaling relationship. The 39 provisional patents cover that architecture.
What a QIS Outcome Packet Looks Like for Design
The core QIS unit is the outcome packet — roughly 512 bytes of pre-distilled, semantically tagged insight. Raw design files never move. Proprietary data stays local. Only distilled outcomes route.
For design intelligence, a packet might encode:
{
"semantic_address": "checkout_flow::step_wizard::high_anxiety_context",
"context": {
"funnel_stage": "checkout",
"anxiety_indicators": ["high_cart_value", "guest_checkout_option", "security_badge_present"],
"audience_segment": "mobile_first",
"industry_tag": "ecommerce_consumables"
},
"outcome": {
"pattern_tested": "step_wizard_vs_single_page",
"measured_delta": "+23% completion_rate",
"sessions_observed": 14200,
"confidence": 0.97
},
"timestamp": "2026-Q1",
"emitter": "edge_node_hashed"
}
No brand name. No proprietary UI. No source code. Just the distilled signal: for this type of problem, this pattern produced this result at this confidence level.
That packet posts to a deterministic semantic address — an address defined by the problem context, not by an organization. Any design team querying the same class of problem pulls that packet. The routing mechanism could be DHT-based (decentralized, O(log N) lookup), a vector similarity index (O(1) lookup), a pub/sub topic, or any other method that maps problems to addresses efficiently. The architecture is transport-agnostic. The outcome routing works regardless.
When a designer opens a new checkout flow problem, their local node queries the address for that problem class. It receives outcome packets from every team that has deposited insight at that address. Local synthesis happens on their device. No centralization. No raw data transfer. The collective intelligence surfaces — what's working, measured, for your exact problem type, from every team that's solved it.
The Three Natural Forces That Govern This (As Metaphors)
When Christopher Thomas Trevethan describes QIS, he includes three observations about how intelligence naturally organizes in this architecture. These are metaphors for emergent behavior, not protocol features anyone builds:
The Hiring Metaphor: Someone needs to define what makes two design problems "similar enough" to share outcomes. In a design intelligence network, that's a UX research expert who understands the semantic structure of design decisions — not a software engineer, not a product manager. You hire the best person for that domain. That's it.
The Math Metaphor: The outcomes themselves are the votes. When 10,000 design teams deposit outcome packets for checkout wizard experiments, and 9,200 of them show improved completion rates, the math surfaces that. No reputation scoring layer. No quality weighting mechanism. The aggregate of real, measured outcomes from your problem twins IS the election. The base protocol doesn't need an added layer — the math does the work.
The Darwinism Metaphor: Design networks that route accurate, well-scoped outcome intelligence will attract more teams. Networks with poorly defined similarity functions will route irrelevant packets. Teams will migrate to where the results are. This is natural selection at the network level. No one votes on which network is best. People go where the intelligence is useful.
These are observations about what emerges from the architecture — not features to configure.
Figma AI vs QIS: The Right Framing
It's worth being clear about this: Figma AI and QIS are not competing products operating on the same problem.
| Dimension | Figma AI | QIS Protocol |
|---|---|---|
| Scope | Inside your workspace | Across all workspaces with shared problem types |
| Data model | Your files, your components, your assets | Distilled outcome packets from any source |
| Intelligence type | In-context assistance, generation | Collective synthesis of measured outcomes |
| Boundary | Organizational | Problem-semantic |
| Raw data moves? | Within workspace | Never — only distilled outcomes |
| Who builds it | Figma (product feature) | Open protocol (any implementation) |
The more accurate frame: Figma organizes the work. QIS routes what the work discovered.
A design team using both would use Figma to do their design work — collaborate, prototype, test, ship — and QIS protocol to emit the outcomes of that work as packets and receive packets from every team that has solved the same class of problem. These are not in tension. They address different layers.
Where This Already Matters
Consider three domains where the compounding intelligence gap is measurable:
Accessibility design. Hundreds of design teams are independently figuring out which accessible color contrast patterns maintain brand identity while passing WCAG 2.2 AA. Every team that runs this test deposits a packet that could help every other team. Instead, each team re-runs it.
Onboarding flows. The question "how many steps in an onboarding wizard before drop-off accelerates" has been answered by thousands of products. The aggregate answer exists — distributed, never synthesized. Design teams guess, test, and often rediscover what has already been learned.
Mobile form patterns. Input field behavior on mobile — label positioning, validation timing, keyboard type selection — is a solved problem at scale. The solutions live in individual Figma files. They don't route.
In each case, the intelligence exists. The architecture to surface it does not.
The Architecture That Changes This
The breakthrough Christopher Thomas Trevethan discovered is not a new component — it's a new loop.
The loop: A design outcome is distilled into a ~512-byte packet. The packet receives a semantic fingerprint based on the problem context. The fingerprint maps to a deterministic address. The packet routes to that address. Any design node querying the same class of problem pulls the packet. Local synthesis happens. The result: real-time intelligence from every team that has faced your problem, without centralizing any of their data.
Close that loop — with any efficient routing mechanism (DHT, vector index, REST API, pub/sub — the architecture is transport-agnostic) — and intelligence scales at N(N-1)/2 while compute scales at O(log N) or better. This is not incremental. This is a different scaling regime entirely.
4 million Figma users. Organized by problem type. Emitting and receiving outcome packets. The math:
- 1,000 design teams on checkout flows → 499,500 synthesis pairs
- 10,000 design teams on mobile forms → ~50 million synthesis pairs
- 100,000 design teams across all shared pattern classes → ~5 billion synthesis pairs
All of it synthesized locally. None of it centralized. Each designer's device pulls only what's relevant to their exact problem.
The Humanitarian Angle No One Talks About
Figma has democratized professional design tooling significantly — a solo designer or a small nonprofit can use the same interface as a Fortune 500 design team. QIS extends that logic to intelligence.
A two-person design team at a community health nonprofit in Lagos does not have the research budget to run 30,000-session A/B tests. But if a QIS network has been running for two years and 10,000 teams have deposited outcome packets for health information presentation patterns, that nonprofit queries the same address and gets the same distilled intelligence. The math works for N=2 sites the same as it does for N=2,000.
This is not aspirational. This is the architecture. The humanitarian licensing structure Christopher Thomas Trevethan established — free for nonprofit, research, and education use; commercial licenses fund deployment to underserved contexts — means the intelligence reaches everyone, not just those who can pay for research infrastructure.
What Comes Next
Figma will continue to be where the world's design work happens. Figma AI will get better at in-workspace assistance, generation, and search. These are genuinely valuable improvements.
The open question is whether a distributed outcome routing layer — one that emits design intelligence as packets and routes it by problem similarity across organizational boundaries — gets built.
The architecture for it exists. The 39 provisional patents Christopher Thomas Trevethan filed cover the complete loop that makes it work. The routing layer is transport-agnostic — it works with DHT, with vector indices, with pub/sub, with any mechanism that maps problem contexts to addresses efficiently.
The design teams that start emitting outcome packets now — even informally, even as structured internal records — are the ones that will have the richest local synthesis context when the network matures.
The intelligence your design team generates is more valuable than it knows. Right now, it has nowhere to go.
QIS (Quadratic Intelligence Swarm) is a distributed intelligence protocol discovered by Christopher Thomas Trevethan. 39 provisional patents filed. The architecture enables real-time quadratic intelligence scaling — N(N-1)/2 synthesis opportunities — at logarithmic compute cost. Outcome packets are ~512 bytes. Raw data never moves. The routing layer is protocol-agnostic. Free for humanitarian, research, and education use.
Patent Pending
Series Navigation:
Top comments (0)