In this PC Mag article, the author begins with a blunt confession: clothes shopping is a chore he actively avoids. That changed when he learned he would be moderating a panel on AI in enterprise at CES. Standing on stage beside sharply dressed executives made his usual uniform of jeans and button-down shirts feel inadequate. He needed to look more polished without enduring the fatigue of wandering through stores.
Instead of turning to a personal shopper, he turned to Google Gemini 3 Pro. Using Gemini’s multimodal abilities, along with its Nano Banana image model, he uploaded casual mirror selfies and asked the system to generate edited versions of himself wearing different outfits. The results were not generic mockups but realistic images that preserved lighting, proportions, and his own appearance. That fidelity allowed him to virtually try on blazers, vests, sweaters, and tweed layers without leaving home.
He refined his prompts over several iterations, requesting variations and receiving not only images but also written analysis of which looks best suited to a professional business setting. This feedback loop helped him narrow his preferences quickly. The process, he notes, would once have required juggling multiple tools and far more effort. Now it unfolded in a single conversational flow.
Armed with clearer intent, he visited a store only once, asked for specific items, and tried on far fewer pieces than usual. While he acknowledges the limits of AI in judging fabric feel, fit, and texture, the preparation made in-person shopping more efficient. After purchasing a blazer, he returned to Gemini to test different shirt colors through photo edits, again using visuals to guide decisions.
By the time he arrived at CES, his wardrobe was sorted and his confidence noticeably higher. The experiment illustrates a broader point. AI did not replace his taste or make choices for him. It acted as a thinking partner and visualization tool, reflecting his intent and expanding options while leaving control in his hands.