By Key Points
- Figma’s AI tool was disabled for having duplicated Apple’s weather app UI
- CEO admits to pushing for a rapid launch without adequate checks
- Unclear training data raises legal and ethical concerns
- AI tools in creative fields must balance innovation with caution
Recently, Figma which is a popular UI design platform faced criticism after its Make Design tool was found to replicate the user interface (UI) of Apple’s weather app.
This incident highlighted the dangers associated with hasty testing and launching artificial intelligence (AI)-)-driven tools into the market.
Figma AI looks rather heavily trained on existing apps.
This is a “weather app” using the new Make Designs feature and the results are basically Apple’s Weather app (left). Tried three times, same results. https://t.co/Ij20OpPCer pic.twitter.com/psFTV6daVD
— Andy Allen (@asallen) July 1, 2024
A Hasty Rollout That Backfired
Make Design by Figma was meant for users to create UI mockups rapidly using generative AI.
However, when Andy Allen who heads Not Boring Software posted two screenshots — one showing an iOS weather app designed by Figma, and another showing Apple’s app.
It became evident that they were almost identical. Upon this discovery, Dylan Field who is the CEO took down the program and admitted responsibility for urging his team to release it too quickly.
Training Data that Raises Questions
The situation compounded as Kris Rasmussen, CTO at Figma revealed that they did not know everything about how their artificial intelligence had been trained.
The system adopted third-party models such as OpenAI’s GPT-4 and Amazon’s Titan Image Generator G1; hence there are concerns about where these models got their training data from. Although Figma stated no content from community files or app designs within Figma itself were used, this event showed why it is risky to use “off-the-shelf” AI models which lack transparent origins.
Drawing Inspiration vs Infringement
This fiasco brings out clearly how close AIs must get between being inspired by something else and infringing upon the rights of others.
As more creativity gets driven by artificial intelligence, creators need to ensure that their software does not blindly copy existing works since this may lead to legal issues for end-users.
Allen’s advice for designers who work with machine learning outputs should always check and modify them before use serves as a good reminder about these responsibilities.
Caution Moving Forward
Figma wants to re-enable Make Design soon, but it will have additional features that make the outputs more variable and prevent accidental copying better than ever before.
The incident is a strong indication of the need for caution in this area which calls for finding the right balance between innovative approaches and safe practices especially when working on designing tools driven by AI technology which changes rapidly.
