This week OpenAI added virtual try-on to ChatGPT Shopping, letting you see clothes rendered onto a photo of yourself before you buy. Reddit is reportedly building something similar to turn its shopping conversations into actual purchases. One industry outlet framed the whole moment bluntly: it's a race over who wins the "AI shopper". And the appetite is already there — in Southeast Asia, nearly eight in ten shoppers now say they turn to AI for product discovery before they even open a store page.
I started LifeGeek because I got stuck looking for one shirt — not because I was excited about AI. So watching the industry sprint toward "AI shopping" this week, the question I keep coming back to isn't "can a model render clothes on your body convincingly." It's a much older one: what actually stops someone from getting dressed with confidence in the first place?
It's rarely the last step. By the time you're looking at a product photo, you've usually already decided you like something. The actual friction sits earlier — standing in front of a closet full of things you own and still not knowing what to wear, or opening a shopping app with no real starting point and scrolling until you give up. A try-on render solves a problem that shows up after you've found the thing. It does nothing for the harder problem of finding it.
That's the part we built LifeGeek around. You don't fill out a style quiz or build a profile before it's useful. You just ask the question you actually have — "what should I wear to this on Saturday," "I'm packing for four days somewhere that keeps flipping between rain and sun, what actually works" — and it works from there, matching you to real items you can go buy, not generated mockups. The weather-aware part matters more than it sounds: a lot of "what to wear" problems are really "I don't know what this place/day is actually going to throw at me" problems, and most style tools never ask.
We're not trying to compete on who can render the most convincing try-on photo. LifeGeek doesn't handle the purchase either — every item it finds links out to the real eBay listing, because our job is helping you land on the right thing, not taking a cut of the transaction. The product and payment stay exactly where they already are; we're the layer that gets you there faster, with less scrolling and no onboarding tax.
The Bain number is the one I think about most, honestly — four out of five people already defaulting to AI for discovery isn't a future trend, it's already true for a huge share of shoppers today. The open question industry-wide isn't whether people will ask an AI for help getting dressed. It's whether the AI they ask actually understands the real-life context behind the question, or just renders a nicer picture of the same generic recommendation. That's the gap we're building for, one real conversation at a time. If you want to see what that looks like for your own closet, it costs nothing to try it, and the pricing is simple if you decide to stick with it.