TL;DR
A designer at Jane Street reports switching from Figma to Claude AI for most prototyping and design tasks. This shift enhances efficiency and reduces reliance on traditional mockups, but raises questions about collaboration and review processes.
A designer at Jane Street now primarily uses Claude AI instead of Figma for creating prototypes and implementing features, marking a significant change in workflow and design process.
The designer, who previously relied heavily on Figma and documentation, reports that over the past two months, they have shifted to using Claude AI for most prototyping and development tasks. This includes building interactive prototypes, testing features, and even skipping Figma entirely for some projects. The process involves describing the problem, prompting Claude to generate a working prototype, and iterating directly within the codebase. This approach has resulted in faster development cycles and more integrated testing, as prototypes are now directly in the code, allowing immediate feedback and refinement. The shift was driven by improvements in AI models, increased familiarity, and the scope of projects expanding beyond simple tasks to more complex, data-driven prototypes. The designer highlights that Claude provides unlimited iteration, enabling rapid experimentation without the traditional back-and-forth associated with mockups and design docs. However, this new workflow raises questions about review processes, as reviewers now see fully functional features rather than mockups, which could impact collaboration and feedback. The designer emphasizes that prototypes are now viewed as living documents, with the final implementation still owned and refined by engineering teams. They also acknowledge potential downsides, such as reduced creative fluidity and the risk of missing novel ideas due to reliance on AI-generated outcomes.
Why It Matters
This shift signifies a potential transformation in design and development workflows, especially in environments where rapid prototyping and iteration are critical. By using AI to generate working features directly, teams can reduce time spent on intermediate steps like mockups and documentation, accelerating innovation. It also empowers designers to test ideas more independently, potentially democratizing prototyping and reducing bottlenecks. However, it raises questions about collaboration, review, and the preservation of creative processes, which could impact team dynamics and project quality.
AI-Powered Prototyping for Non-Designers with Claude AI: How to Use Claude to Build Product Mockups, Slide Decks, and Brand Materials from Scratch
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Background
Historically, designers relied on tools like Figma for visual mockups and documentation, with prototypes serving as communication tools rather than working code. Over recent years, AI tools have begun to influence workflows, but adoption was limited to small tasks. The recent advances in large language models (LLMs) and AI-generated code have enabled a more integrated approach, allowing designers and developers to prototype directly in code. At Jane Street, this transition has been accelerated by improvements in AI capabilities, enabling the creation of complex, data-driven prototypes and even skipping traditional design tools for certain projects. The shift reflects broader trends in software development and design, emphasizing rapid iteration and AI-assisted creation.
“Over the past two months, I’ve shifted from Figma to using Claude AI for most prototyping and feature development.”
— Jane Street designer
“Claude provides unlimited iteration, enabling rapid experimentation without the traditional back-and-forth of mockups and docs.”
— Jane Street designer
“Prototypes are now living documents, with the final implementation still owned and refined by engineering teams.”
— Jane Street designer
Designing and Prototyping Interfaces with Figma: Elevate your design craft with UX/UI principles and create interactive prototypes
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What Remains Unclear
It is still unclear how widespread this workflow will become across teams, how review processes will adapt long-term, and whether reliance on AI might limit creative exploration or lead to overlooked ideas. The long-term impact on collaboration dynamics and project quality remains to be seen as the practice evolves.
KI-gestütztes Prototyping für Nicht-Designer: Wie man mit Claude Produkt-Mockups, Präsentationen und Markenmaterialien von Grund auf erstellt (German Edition)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What’s Next
Next steps include developing formal guidelines for review and collaboration in AI-driven prototyping, assessing the impact on team workflows, and expanding the use of AI tools for larger and more complex projects. Monitoring how this approach influences project outcomes and team dynamics will be key.
Version Control for Web Developers: Building Collaborative and Scalable Application Workflows
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
Why is the designer shifting from Figma to Claude AI?
The designer reports that AI-generated prototypes allow for faster iteration, reduce the need for intermediate mockups, and enable direct testing of features within the codebase, improving efficiency.
Does this change affect collaboration and review processes?
Yes, prototypes now function as fully baked features, which may alter how feedback is given. The team is treating these prototypes as living documents, with final ownership still in engineering, but the process is still being refined.
Are there risks associated with relying on AI for design and prototyping?
Potential risks include reduced creative fluidity, missing novel ideas due to AI’s limitations, and challenges in collaboration. The team is aware of these issues and is experimenting with workflows to address them.
Will this workflow be adopted across all projects?
It is currently a developing practice within one team at Jane Street. Broader adoption depends on further evaluation of its effectiveness, collaboration impact, and project complexity.
Source: Hacker News
