Why It Matters
When a co‑founder of a design‑first unicorn questions the foundation of a protocol that powers AI‑driven collaboration, the ripple effects reach beyond the immediate product. Dylan Field’s recent remarks about the Multi‑Channel Protocol (MCP) aren’t just a technical gripe; they hint at a broader tension between openness, scalability, and the creative control that designers crave.
Context and the MCP Debate
Figma, the cloud‑based design platform that reshaped how teams iterate on UI/UX, has long advocated for interoperable standards that let AI assistants stitch into design workflows. MCP emerged as a joint effort among several AI labs to create a common language for exchanging prompts, model outputs, and context across tools. While the ambition was noble, early adopters have reported friction points—latency spikes, inconsistent metadata handling, and limited extensibility.
According to TechCrunch, Field observed that “MCP isn’t perfect.” This brief comment encapsulates a growing sentiment: the protocol works, but it falls short of the fluid, real‑time experience designers expect when they toggle between sketching, prototyping, and AI‑generated suggestions.
Field’s Critique Explained
Field’s concerns can be grouped into three practical dimensions:
- Latency and Responsiveness: Designers need sub‑second feedback. MCP’s batch‑oriented design introduces noticeable lag, especially when large language models process complex prompts.
- Metadata Fidelity: Design assets carry layers of intent—style guides, accessibility notes, brand constraints. The current protocol often strips or simplifies this metadata, forcing designers to re‑enter constraints manually.
- Extensibility for Niche Tools: Smaller plugins and emerging AI‑assistants struggle to map their proprietary data structures onto MCP, limiting innovation at the edges of the ecosystem.
What This Means for the Industry
The criticism signals that the design‑tech community may gravitate toward more bespoke, lightweight standards tailored to creative workflows rather than a one‑size‑fits‑all AI messaging layer. Companies like Adobe and Canva have already hinted at internal “design‑first” protocols that prioritize immediate UI feedback and richer semantic tags.
Moreover, Field’s public acknowledgment could accelerate open‑source contributions to MCP, pushing its governance model toward a more community‑driven approach. If the protocol evolves to address latency and metadata loss, it could become a true lingua franca for AI‑augmented design, unlocking cross‑tool collaborations that are currently speculative.
Looking Forward
In the short term, developers building on Figma’s platform are likely to implement hybrid solutions—layering MCP with proprietary shortcuts to keep the user experience snappy. Longer‑term, we may see a divergence where “design‑centric” protocols coexist alongside broader AI messaging standards, each serving distinct user personas.
Field’s candidness also underscores a cultural shift: designers are no longer passive recipients of AI output; they demand agency, transparency, and performance. As AI continues to embed itself into creative pipelines, the protocols that mediate this relationship will become as critical as the models themselves. Watching how MCP adapts—or is superseded—will be a bellwether for the health of open collaboration in the next wave of design intelligence.
Original reporting via Source.