Key Takeaways:
Figma’s new tool, “Make Designs,” which allows users to mock up apps using generative AI quickly, has been pulled after the tool drafted designs that looked strikingly similar to Apple’s iOS weather app.
Figma CEO Dylan Field posted a thread on X early Tuesday morning detailing the removal, taking personal responsibility for pushing the team to meet a deadline and defending the company’s approach to developing its AI tools.
(1) As we shared at Config last week – as well as on our blog, our website, and many other touchpoints – the Make Design feature is not trained on Figma content, community files or app designs. In other words, the accusations around data training in this tweet are false. https://t.co/jlfmroPPhm
— Dylan Field (@zoink) July 2, 2024
Andy Allen, CEO of NotBoring Software, revealed that Figma’s Make Designs tool made near-replicas of Apple’s weather app. He advised designers using the feature to thoroughly check existing apps or modify the results heavily to avoid legal trouble.
Figma’s Response
In an interview, Figma CTO Kris Rasmussen stated that Make Designs was not trained on Apple’s app designs and that Figma did not train the AI models it used.
Rasmussen noted, “The features are powered by off-the-shelf models and a bespoke design system that we commissioned, which appears to be the underlying issue.” Field reiterated that Make Designs is not trained on Figma content, community files, or app designs, emphasizing, “the accusations around data training in this tweet are false.”
He attributed the problem to the low variability in the design system.
Technical and Policy Insights
The key AI models powering Make Designs are OpenAI’s GPT-4o and Amazon’s Titan Image Generator G1. If these models were responsible for the similarities to Apple’s designs, it suggests that the third-party models might have been trained on such designs, although this remains speculative.
Haha wow. Can Figma AI only produce one screen at a time or could you ask it to lay out a multi-screen flow?
— Andrew Baker (@andr3wbak3r) July 1, 2024
Figma introduced new AI training policies last week, allowing users until August 15th to opt in or out of having their content used for Figma’s training.
Rasmussen highlighted that Figma plans to eventually train its own models to perfect design workflows while ensuring only general design patterns and specific Figma concepts are learned.
Addressing the Issues
Rasmussen assured that Figma is reviewing its bespoke design system to ensure sufficient variation and adherence to quality standards.
I mean. Most UI/UX is heavily based on R&D done of what “works” when predicting user intention. Most startups base their protos and further on “existing apps” anyways.
— PEXAVC (@pexavc) July 3, 2024
The Make Designs feature will remain disabled until these issues are addressed. Despite the controversy, other Figma AI features will continue to be available in beta. Users interested in these features can sign up for a waitlist.
Industry Impact and Ethical Considerations
Figma is the latest in a series of companies facing scrutiny over AI integration in creative tools. This includes Adobe, which had to clarify its AI training policies, and Meta, which had to adjust its AI labels following user complaints.
I mean, that IS probably the most common/used weather app…
— matty (@allthegweiup) July 2, 2024
The controversy underscores the delicate balance between leveraging AI for efficiency and maintaining originality and ethical standards in design. Figma’s swift action to disable the Make Designs feature highlights the company’s commitment to quality and ethical standards.
Surprise surprise. AI copies more shit
— JJ Ferman (@jjferman) July 2, 2024
As the debate over AI-generated content continues, Figma’s approach to transparency and rigorous testing will be crucial in navigating these complex issues.
For more news and trends, visit AI News on our website.
