The New York subway has become an unlikely gallery of AI-era creativity.
There are panties that think. There is a personal injury lawyer riding a giant pigeon over Manhattan. The images are absurd, arresting, and difficult to ignore. Whether they represent brilliant creative thinking or simply AI-generated excess, they raise an uncomfortable question: when almost anything can be made to look like something, what makes an idea worth making?
For design researcher and creative strategist Rhea Soman, this question goes beyond the novelty of AI-generated imagery. It touches on a more fundamental shift in how people create, interact with technology, and assign meaning to the things they make.
Much of Soman’s perspective comes from seven years working with Lumière Business Solutions, a technology-forward research and consulting agency. It was here that she developed her understanding of the relationship between design and research, working across disciplines to translate human insights into tangible ideas, experiences, and strategies. Her work involved research-driven projects across industries, including collaborations with global organizations such as Unilever.
This experience shaped how she approaches creative problems: not simply by asking what something should look and feel like, but by investigating the people, behaviours, and motivations that shape it. For Soman, research is not separate from the creative process. It is what gives creative decisions their direction and purpose.
Her interest in the relationship between technology and creativity also extends into her own speculative design practice. In Echo & Eros, her speculative magazine project, Soman explored artificial intelligence through the lens of beauty, using fictional brands and imagined futures to question society’s growing obsession with manufacturing the ideal version of ourselves. Drawing on ideas explored in Jean Baudrillard’s Simulacra and Simulation, the project considered what happens when representations stop merely reflecting reality and begin shaping it.
Generative AI makes that question newly tangible. It can reproduce the visual language of almost anything, from luxury advertising to artistic experimentation. But when the appearance of creativity becomes easy to reproduce, what distinguishes an image that looks creative from an idea that means something?
Soman’s subsequent work with AI creative platform Virse brought her to a more practical version of the same question: what happens when technology starts shaping the process of making things?
After experimenting with tools such as Magnific, Adobe Firefly, and Replicate, she found herself moving between platforms to develop ideas. Generating an image was relatively straightforward. Keeping track of the thinking behind it was another story.
Through research and creative strategy work with Virse, she began exploring how creatives understand AI, where their workflows break down, and what they need beyond the ability to generate.
Understanding where technology fits
One of the clearest themes emerging from Soman’s research with Virse was the gap between what a product can do and how people understand it.
In a workshop designed to explore creative workflows and perceptions of AI, participants imagined possible applications for the platform, but its role was not always immediately clear. The challenge was not simply explaining features. It was helping people understand where the product belonged within the way they already worked.
It is a familiar design problem: a platform can have impressive capabilities, but if users need a decoder ring to understand its purpose, the experience has already created friction.
For Soman, this is where design research becomes particularly valuable. It shifts the question from “What can we build?” to “What do people need to understand before they can use it?”
Rather than leading with AI terminology or a list of technical capabilities, the product needed to communicate a recognizable purpose and demonstrate how it could support existing creative practices.
Keeping the human in the creative process
Another theme that surfaced was the anxiety surrounding AI’s potential to replace designers. Yet the conversations also revealed a counterpoint: creatives were interested in using AI to reduce friction, not surrender authorship.
Designers make decisions constantly. They choose directions, reject alternatives, refine details, and recognize when something finally feels right. Creative work depends on judgment, and that judgment cannot be reduced to generating more options.
AI can accelerate parts of this process, but speed is not the same as creative freedom. Endless variations are not particularly useful if someone is left sorting through them without getting closer to an answer.
This distinction influenced Soman’s thinking around Virse’s positioning. Instead of presenting AI as the creative hero, the brand needed to communicate a human-led partnership. Virse connects the work; the designer directs it.
The ambition was not to automate creativity out of existence, but to remove enough friction that people have more room to exercise it.
The problem with fragmented creative workflows
Creative work rarely happens in isolated steps. Ideas move between references, assets, tools, conversations, feedback, and revisions. When those pieces become scattered, context gets lost and momentum suffers.
Soman’s own workflow reflected this fragmentation. Each platform offered something useful, but moving between them often meant transferring files, reconstructing decisions, and trying to preserve the thinking behind an idea.
The Virse research explored concepts such as iteration, refinement, collaboration, context, and progress. Together, these pointed toward a larger opportunity: supporting the journey of an idea rather than simply generating its next output.
Virse’s one-platform approach responds to this fragmentation by bringing creative context together. Its ambition is to help people move from exploration to refinement without repeatedly reconstructing the thinking behind their work.
But Soman’s research also points to an important distinction: putting tools in one place does not automatically create continuity. A useful platform must help people preserve intent, build on previous decisions, and remain in control of the creative process.
The real benefit is not having more tools at one’s fingertips. It is spending less energy managing them and more energy creating.
What research across industries can teach AI companies about trust
An unexpected connection emerged when Soman began interpreting the research through the lens of her earlier work in banking and healthcare.
Across these sectors, she had encountered recurring questions around trust, onboarding, loyalty, and retention. The industries were different, but the underlying challenge was familiar: how do people learn to use a product, what makes them confident in it, and what gives them a reason to return?
The same questions apply to AI creative tools.
Curiosity is often the easy part. A striking result can make someone fall in love with a tool in minutes. But will they trust it with a real project, a client deadline, or an idea they have spent weeks developing?
Loyalty presents another challenge. The digital landscape is full of products people try once and abandon. For an AI tool to earn a lasting place in someone’s workflow, it needs to offer value beyond the first impressive result.
These questions also informed Soman’s work on Virse’s brand strategy, visual identity, and communications. Her work explored different ways people might understand the product, from an AI operating system to a collaborative creative workspace and a professional design partner.
The resulting direction emphasized shared context and the ability to scale creative work without replacing human judgment.
The brand needed to make that promise tangible. Its identity communicates connection and movement, while its voice aims to be human, calm, and curious rather than intimidating or filled with technology hype.
For Soman, branding in this context is not simply about looking distinctive. It is about setting the right expectations for the experience people are about to have.
Beyond the prompt
Soman’s work with Virse reinforces a distinction at the center of the current conversation about artificial intelligence: generating something and creating something are not necessarily the same thing.
AI may make it easier to produce almost anything. But helping people understand a tool, preserve their creative judgment, carry ideas forward, and build confidence in the experience requires a different kind of design thinking.
Her work suggests that the next frontier of AI creativity may not be simply generating more, or even generating better. It may be about building tools that help ideas move forward without taking the human out of the process.
After all, the goal was never to spend more time switching tabs. It was to make something worth making.
About Rhea Soman
Rhea Soman is a multidisciplinary designer, researcher, and creative strategist based in New York. With over eight years of industry experience, she has worked across technology, consumer goods, healthcare, and financial services. She spent seven years with Lumière Business Solutions, a research and consulting agency, exploring the intersection of human behaviour, technology, and design through projects for global organizations. Her recent work includes AI creative platform Virse, where she contributed to research, brand strategy, visual identity, and communications, as well as SENA, a nonprofit focused on women’s safety in New York City, where she contributes to strategy and design research. She holds a postgraduate degree in Communication Design from Pratt Institute.
