Get your free personalized podcast brief

We scan new podcasts and send you the top 5 insights daily.

Instead of a traditional design handoff, Eleven Labs built custom WebGL playgrounds for its "voice orb" project. This allowed designers to manipulate shader parameters with simple controls, take screenshots of desired aesthetics, and feed them back to the engineer, blending creative and technical workflows.

Related Insights

Standard AI coding tools force a linear A-to-B iteration process, which stifles the divergent thinking essential for design exploration. Tools with a 'canvas' feature allow designers to visualize, track, and branch off multiple design paths simultaneously, better mirroring the creative process.

Contrary to traditional digital design, the modern AI-assisted workflow involves broad, conceptual exploration on canvas-like tools (e.g., Paper) and sweating the final visual details directly in code. Pixel-nudging in design software like Figma is becoming obsolete for last-mile fit and finish.

To solve the problem of representing a non-visual product, Eleven Labs created a generative system. It uses the underlying data embedding of each voice to drive a real-time WebGL shader, creating a unique, visually representative "orb" for each of the 40 million voices on its platform.

Instead of being limited by off-the-shelf software, designers can dramatically accelerate their process by building bespoke tools. MDS used the AI tool V0 to create a custom bitmap icon builder, enabling rapid prototyping of a unique interactive element.

Mature AI design workflows involve spending significantly more time in low-cost, exploratory environments like HTML prototypes before a full build. This "visual planning" phase solidifies UX and flows, proving more efficient than the earlier method of immediately generating and then tweaking code.

Manually creating design variations is slow. Instead, build a simple internal tool with sliders to control parameters like wave functions, colors, and spacing. This "parametric visualization" allows for rapid, real-time exploration of a massive design space, leading to more unexpected outcomes.

Creating custom "playground" tools for design exploration no longer requires advanced coding. You can simply describe the interface and the controls you want (e.g., "a grid with sliders for rows and opacity") in a natural language prompt to an AI, which will generate a functional tool.

Instead of manually tweaking complex parameters like animation timing or color variables, designers can now ask an AI to build a custom visual editor for that specific task. This approach creates a hyper-efficient, temporary tool that provides precise control and can be discarded after use, a process that was previously unjustifiable.

At Eleven Labs, product design isn't about applying existing patterns. When the research team develops a new model with unprecedented capabilities, the designer's job is to invent a completely new interface. This requires the ability to discard classical UI conventions and imagine interactions from scratch.

When exploring an interactive effect, designer MDS built a custom tool to generate bitmap icons and test hover animations. This "tool-making" mindset—creating sliders and controls for variables—accelerates creative exploration far more effectively than manually tweaking code for each iteration.