The 'Clueless' Closet Is Real, But Will It Actually Make You Stylish?
Alta raised $11M to bring AI virtual try-on tech to fashion. With 100M+ outfits generated, the question isn't whether the tech works—it's whether algorithms can truly understand style.
100 million outfits have been digitally tried on since 2023. That's roughly one outfit for every three Americans, all generated by AI avatars that never get tired of changing clothes.
Jenny Wang's Alta has turned the fantasy from "Clueless" into reality, complete with $11 million in funding led by Menlo Ventures. But as the company debuts its first brand integration with Public School during New York Fashion Week, a bigger question emerges: Can technology truly understand style?
Speed Kills the Competition
While major brands like Zara have dabbled in digital avatars, Alta's edge is raw speed. Zara's avatars take two minutes to try on four items. Alta's can style eight pieces in seconds. In the instant-gratification economy, those seconds matter.
Public School's integration marks a shift. Instead of forcing users to leave a brand's website, Alta embeds directly into the shopping experience. "It's not 2015 anymore," said Public School's Dao-Yi Chow. "We have to look at tech as a partner."
The Data Dilemma
Alta's ultimate vision goes beyond virtual try-ons. Wang wants to create the "personal identity layer for the future of consumer AI." Translation: Your avatar becomes your shopping agent, knowing your body, preferences, and purchase history better than you do.
That's where things get complicated. Building this "identity layer" requires intimate data about users' bodies, shopping habits, and style preferences. In an era of increasing privacy concerns, will consumers trust a startup with their digital identity?
The company's investor list—including models Jasmine Tookes and Karlie Kloss—suggests fashion insiders believe in the tech. But fashion expertise and consumer adoption are different games.
The Algorithm vs. Authenticity Problem
Here's the paradox: Fashion is deeply personal, yet Alta's business model depends on scalable algorithms. Can AI truly capture the intangible elements that make an outfit work—confidence, mood, cultural context, or that inexplicable "je ne sais quoi"?
Consider this: Alta has generated 100 million outfits, but how many were actually purchased and worn? The company hasn't shared conversion metrics, which might be the more telling statistic.
What Retailers Really Want
For brands, virtual try-on solves a different problem than consumer styling. It reduces returns, increases engagement, and provides rich data about customer preferences. Public School's integration isn't just about helping customers—it's about learning from them.
The real test will come when Alta expands beyond early adopters to mainstream consumers who might find AI styling overwhelming rather than helpful.
Authors
Related Articles
Samsung Electronics' union branch revealed that 84% of survey respondents oppose the government's roughly $290 billion (₩400 trillion) semiconductor megaproject in Gwangju. Government, management, the union and shareholders are all reading the same national project in sharply different ways.
Nvidia shipped roughly a billion RISC-V cores in 2024, then announced it would run CUDA on the open standard. We break down how royalty-free instruction sets and open software stacks are trying to route around CUDA's lock-in. Part 2 of the Semiconductor Sovereignty series.
US AI-chip export controls split into three layers in the first half of 2026 — January easing, a May crackdown on circumvention, and a pending bill. Nvidia erased China from its guidance and still posted a record $81.6 billion quarter. A look at the export policy that both shields and cages it.
AMD's MI325X matches or beats Nvidia on memory and bandwidth — yet Nvidia's 86-92% share holds. The real moat is CUDA, 20 years in the making. Part 1 of 4.
Samsung Electronics' union branch revealed that 84% of survey respondents oppose the government's roughly $290 billion (₩400 trillion) semiconductor megaproject in Gwangju. Government, management, the union and shareholders are all reading the same national project in sharply different ways.
Nvidia shipped roughly a billion RISC-V cores in 2024, then announced it would run CUDA on the open standard. We break down how royalty-free instruction sets and open software stacks are trying to route around CUDA's lock-in. Part 2 of the Semiconductor Sovereignty series.
US AI-chip export controls split into three layers in the first half of 2026 — January easing, a May crackdown on circumvention, and a pending bill. Nvidia erased China from its guidance and still posted a record $81.6 billion quarter. A look at the export policy that both shields and cages it.
AMD's MI325X matches or beats Nvidia on memory and bandwidth — yet Nvidia's 86-92% share holds. The real moat is CUDA, 20 years in the making. Part 1 of 4.
Thoughts
Share your thoughts on this article
Sign in to join the conversation