Weber Wong Wouldn’t Have Backed Himself. Then He Went to Art School.

Weber Wong had spent enough time looking at founders to reach an awkward conclusion about himself. He liked venture capital. He was learning. He could recognize an interesting business. But when he imagined sitting across the table from his own pitch, he could not quite picture writing the cheque. The verdict came with an unusually direct remedy: leave the job, move to New York, make some art.

There was a coffee shop in the next chapter. Then there was graduate school. Eventually there was a camera strapped to his head, broadcasting his life through artificial intelligence for two weeks. For someone who had worked in investment banking and venture capital, this was an admirably difficult career move to explain in a spreadsheet.

It also led to FLORA, the creative software company he cofounded with Alex Li. By January 2026, the company was announcing a $42 million Series A. The detour had produced something Wong had struggled to find while examining other people’s companies: work he wanted to spend his own life doing.

An investment committee of one

Wong’s career passed through Evercore and Menlo Ventures before reaching NYU’s Interactive Telecommunications Program, usually called ITP. Those first two names belong comfortably together on a finance CV. The third opens onto a room where technology is a medium for making art.

His interest in art preceded the startup. Poetry and history appear among the interests he discusses, alongside the mechanics of building a company. The combination matters because the familiar finance-to-founder story tends to make the second job sound like the inevitable promotion. Wong’s account is messier. He wanted to make things, and the companies he encountered as an investor did not necessarily resemble things he wanted to make.

He could appreciate their prospects without feeling personally enlisted in their missions. A promising business can still be somebody else’s business. Recognizing that distinction is easy in retrospect, once a founder has a company and a funding announcement. Acting on it before having either requires tolerating a less impressive answer to the question of what you do.

“I wouldn’t have backed myself at the time.”

Weber Wong

The art-school route back to technology

In New York, Wong worked at a coffee shop and explored the art scene. ITP offered a way to bring his interests together. He had encountered the program through Runway, whose founders had studied there. At ITP, computation could serve an installation, an interaction or an experiment. The technology did not have to begin with a market.

That was a useful reversal for a former investor. Rather than starting with a category of companies, he could start with an object he wanted to make. Rather than ask whether a market was attractive, he could ask why a particular creative experiment was so awkward to assemble.

His projects with Li involved real-time AI installations. In his early public announcement, Wong described the frustration plainly: advanced techniques were available, but access to them often depended on engineering expertise. An artist might know what an experience should feel like while lacking a convenient way to direct the technology producing it.

There is a particular sort of software idea hidden in that frustration. A tool can be technically available and practically out of reach. Documentation helps. So does a patient technical collaborator. Neither quite substitutes for an interface that lets the person making the work remain in charge of it.

Please hold. The founder is broadcasting.

The early demonstration was wonderfully literal. Wong wore a GoPro and streamed his life, letting people see the world around him transformed with real-time AI. He committed two weeks to it. A landing page became a place to watch the technology work on an actual person’s surroundings.

Most product demonstrations have the good manners to end after a few minutes. This one followed the founder around. It suited the project’s origins: an interactive artwork could also show what the underlying tools made possible. The spectacle invited attention, while the machinery behind it suggested something visitors might use themselves.

His July 2024 announcement connected the company explicitly to the installations he was making with Li. He also named investors, including his former firm, Menlo Ventures. Art and business were already sharing the same page.

The livestream is a memorable detail because it puts Wong inside the experiment. He was willing to be the raw material. Before FLORA became an environment for other people’s creative processes, its introduction involved making his own daily experience available for transformation.

The second decision is the interesting one

FLORA’s visual canvas gives the argument a physical shape. Text, images and video can occupy connected steps. An output can become an input. A direction can branch. The developing process stays available to inspect, rather than vanishing behind the latest result.

Consider a creative team discussing a campaign. Producing an image might be the first task. Choosing among several directions, preserving the useful references and deciding what to change are further tasks. A room full of attractive options can still leave everyone uncertain about the next move. The software has to help with that part too.

Wong’s manifesto pairs creative ease with creative control. It also explains the company’s name through an aspiration that creation should feel natural. There is a small irony here: a canvas of connected boxes can look decidedly technical. The ambition is for that structure to become useful enough that the person working stops thinking about the structure.

The difference is visible in the questions a professional asks. Can this detail change while the rest stays useful? Can a colleague understand the path to this version? Can the team repeat the process? These are ordinary production concerns. They become especially pressing when a model is capable of delivering an unexpected answer every time.

The old firm returns, this time with a cheque

On May 28, 2025, FLORA announced a $6.5 million seed round led by Mike Volpi at Hanabi Capital. Menlo participated, alongside investors including Justin Kan and Gabe Whaley. Wong’s former professional world had returned in a new capacity. He was now bringing it a creative tool.

The next January, a $42 million Series A led by Redpoint Ventures brought announced total funding to $52 million. These figures describe financing, rather than sales or personal wealth. Their practical significance was the room they gave the company to build and help teams adopt the product.

Wong’s funding essay described teams using FLORA at Pentagram, AKQA, Red Antler, Lionsgate and MSCHF. It also introduced a deliberately human part of the operation: Forward Deployed Creatives, who work directly with teams to help fit generative tools into their processes.

That job title carries a useful admission. Even software intended to make creation feel natural can need someone nearby to explain it. Habits, handoffs and shared judgment are harder to install than an application.

“We treat models as steps in your creative process, not destinations.”

Weber Wong, January 2026

Someone still has to know what good looks like

Wong’s advice about delegation starts with doing the work yourself for as long as you can manage it. Learn the task well enough to recognize a good result before handing it to someone else. It is a practical rule for an expanding company, and it sounds consistent with a founder who initially built tools for his own projects.

The rule also helps explain his approach to customers. Creative work contains judgments that are difficult to express as a feature request. Sitting with a team, watching where it gets stuck and seeing what it accepts as finished can teach something a polished demonstration misses.

In March 2026, Wong introduced FAUNA, a creative agent within FLORA. The announcement emphasized that the steps it takes remain on the canvas. Prompts, models and sequences stay visible after a session. Delegation, in this version of the product, comes with a record of what happened.

That matters for the next person opening the work. A result can impress while leaving a colleague unable to reproduce or revise it. Keeping the process available makes the work easier to discuss. It also leaves somewhere for the human author to intervene.

A different entrance to the same canvas

By September 2026, Wong was discussing Fashion Studio: an interface organized around tasks fashion professionals already recognize. The powerful node-based canvas remained, but it was no longer the only way to encounter the work. Familiar activities could become the entrance to the generative machinery underneath.

He acknowledged the learning required by nodes, model selection and prompting. He also discussed the limits of precision in generated imagery. A concept can be useful before it is reliable enough for every production use. Those distinctions matter in a field where an appealing picture and a producible garment have different requirements.

The development brings his story back to its original problem. At ITP, Wong wanted access to techniques without having to surrender creative direction. Now he was reconsidering how much machinery a professional should have to understand to use that access.

His early verdict on himself had changed too. Having found the work he wanted to do, he could imagine backing the person doing it. The company began with an investor stepping away from a promising career to make art. Its continuing task is more specific, and less theatrical than a two-week livestream: help someone with an idea make the next deliberate choice.