Dylan Field

Dylan Field is the co-founder of Figma, identified as its chief executive officer in his June 2026 Config announcement. His AI strategy centers on making software easier to create while preserving room for human design judgment. In an October 2025 interview, he described his strategic rule: “Assume AI models get better and make sure that makes Figma better.” He argued that cheaper code generation and more software competition would make brand, taste, craft, and design more consequential.[3][1]

Field sees natural-language prompting as an early interface rather than the endpoint of AI interaction. In October 2025, he predicted that contemporary tools would come to resemble “the MS-DOS era of AI,” with other interfaces allowing people to explore model capabilities more intuitively. His stated aim for Figma was to help users explore aesthetic possibilities rather than impose a personal aesthetic viewpoint. He also argued that involving more people in design would expand designers’ responsibility for guiding exploration and thinking through complete experiences.[1]

Figma Make gives concrete form to that direction. Introduced in May 2025, it could receive existing Figma Design frames with their structure and metadata, then use prompts to turn them into interactive experiences. Users could point to an element and describe a change. The launch announcement, authored by designer Peter Ng and engineers Rohit Chouhan and Tom Duncalf, specified Claude 3.7 Sonnet as the model then in use. Field subsequently described Make as a less intimidating entry point to visual design: users could generate a starting point, copy it into Figma Design, and manipulate it visually.[4][1]

Field distinguishes disposable prototypes from work in large codebases with established patterns. For the latter, he explained Code Connect as a formal mapping between corresponding Figma components and code components, supplying context through MCP so developers can implement designs consistently with existing code. He allowed sources of truth in design and code to coexist: some design libraries mirror implemented components, while others describe work that has yet to be built. In October 2025, he also acknowledged that Code Connect remained too difficult to set up.[1]

Field identified Figma’s edit journal and the history of MCP calls as potential context for agents to understand what had changed between requests. He treated this as an opportunity, not a completed capability: when asked whether an LLM could understand the aesthetic changes in a Figma diff, he said that was not yet available. He likewise emphasized that translating a design into code covers only part of front-end engineering; states, intended behavior, and the complete experience still require work.[1]

In his June 24, 2026 Config article, Field sharpened his distinction between easier creation and greater creative ambition: “But while AI has lowered the floor, it has not raised the ceiling.” He assigned that latter task to designers, creatives, and builders. He also argued that code should be treated as a design material alongside images, vectors, and design layers. His announcement described code layers for exploring interactive alternatives side by side, extracting editable design layers, and updating code after visual changes, with early access scheduled to begin rolling out in July.[3]

Before these AI products, Field and Evan Wallace had pursued browser-based creative tools. Their earlier work supplies context for Figma’s visual and collaborative approach: it involved both making graphics software practical on the web and learning how designers responded to a more open workflow.[6][7]

Field studied computer science and mathematics at Brown University and interned at O’Reilly Media, LinkedIn, and Flipboard before Figma. In his 2015 launch account, he recalled Wallace showing him a WebGL reimplementation of a server-side image-processing API at Brown in April 2011. Over the following year, they discussed making creative tools accessible in the browser and decided to start a company with support from the Thiel Fellowship. Field later recalled exploring meme generation and photo editing before concentrating on interface design.[6][5]

Field’s interest in machine learning also predates Figma’s generative AI products. He recalled working with classical machine-learning approaches during his LinkedIn internship and exploring computational photography with Wallace, explicitly crediting Wallace with most of their early demos. He approximately dated an encounter with Chris Olah demonstrating a small neural network to 2014 or somewhat earlier. Field initially dismissed it as addressing an already solved vision problem, then became more attentive to Olah’s work. He described GPT-3 as a later moment when the improvement over earlier models convinced him something important was happening.[1]

On December 3, 2015, Field announced Figma’s preview release after three years of team work. He described vector rendering, font layout, and performance edge cases as major challenges in building a browser graphics tool. The preview supported collaboration through links, contextual feedback, and shared brand colors; simultaneous multiplayer editing was listed as future work for 2016. In his March 2025 conversation with Garry Tan, Field credited Wallace with developing the real-time collaboration engine after the closed beta.[6][5]

Field’s understanding of collaboration changed after launch. In a December 2020 retrospective, he said he had initially underestimated resistance from designers who feared greater transparency, micromanagement, and loss of control. He came to understand the browser decision as a cultural change as well as a technical one. In his later conversation with Tan, he recalled a public “design party” that began with people mocking Figma and became an example of collective creation; the team spent the following 48 hours working to keep the shared file from crashing.[7][5]

Field describes intuition as “a hypothesis generator that you have to test the hypotheses.” He prefers bringing users’ voices to the team, drawing on support, sales, community conversations, and qualitative and quantitative research. He recalled an early, hours-long user test with Payam, then a designer at Coursera, who followed up with an eight-to-ten-page document anticipating much of Figma’s eventual roadmap. In another early-customer story, Field recalled turning the car around after onboarding Coda at Shishir Mehrotra’s office when an email reported broken fonts; they returned immediately to fix the problem.[1][5]

On December 18, 2023, Field announced that Figma and Adobe had jointly decided to end their pending acquisition. He explained that, after 15 months of regulatory review, the companies no longer saw a path to approval. Field said the agreement had aimed to accelerate what both companies could do for their communities. Continuing independently, he set out a direction of helping anyone design and build digital products on a single multiplayer canvas, from idea to production, and described advances in AI as bringing Figma’s founding imagination-to-reality aspiration closer within reach.[8]