91导航

When David Holz spoke to Regeneron ISEF finalists this past May, he hinted that Midjourney, his generative AI company best known for turning text prompts into images, was pursuing some unexpected new projects.

鈥淲e鈥檙e working on a bunch of hardware projects,鈥� David told the finalists, describing experiments involving sensors, lasers and large-scale machines. 鈥淲e have a lot of really weird projects.鈥�

Those plans have since started coming into focus. This summer, Midjourney with an experimental full-body ultrasound scanner intended to create detailed three-dimensional maps of the body. The company is also planning a that would combine the scanners with baths, saunas and other wellness amenities.

The unexpected venture is the latest chapter in a scientific journey that began long before Midjourney. David was named to the . But before all of that, he competed at ISEF in 2006. He returned two decades later as the founder and CEO of Midjourney, which also sponsors the Technology Enhances the Arts (TECA) category at Regeneron ISEF, and as a member of the 91导航 Board of Trustees for a conversation with Society President and CEO Maya Ajmera.

From ISEF to an entrepreneurial career

David described himself as 鈥渏ust a random kid in Florida鈥� who did not grow up surrounded by scientists. He said every adult who 鈥渃ared and listened and encouraged鈥� had 鈥渁 bigger effect than you鈥檇 think.鈥�

His 2006 ISEF project focused on locating and tracking sounds in three dimensions using microphones. 鈥淢y board had a bunch of microphones built into it and the wires going into a computer,鈥� David recalled. 鈥淎s we would walk around the board and talk about the project, it would show in 3D where we were based off our voices.鈥�

The project reflected an early interest in sensors and spatial tracking that later carried into Leap Motion, the advanced hand-tracking technology company he co-founded. 鈥淚 was really into sensors and tracking and understanding things, and that ended up going into my first company, which was Leap Motion,鈥� David said.

David left his Ph.D. program at the University of North Carolina at Chapel Hill to launch the company. He had originally imagined a more traditional career path: completing his doctorate, working at NASA and eventually running his own lab. Then he began to wonder whether he needed to wait.

鈥淲hat if I just skip a lot of those steps and go straight to trying to make my own thing?鈥� he recalled asking himself. 鈥淪o I was like, 鈥極kay, I鈥檓 going to take six months, 12 months, see what I can do.鈥� And never came back.鈥�

That independent streak also shaped how David built Midjourney, which became profitable without outside investors. For David, earning revenue directly from customers was an important test of whether the product was genuinely liked by customers. 鈥淩aising money is very different than having a product that people are willing to pay for,鈥� he said. 鈥淚n some ways, if people aren鈥檛 willing to pay for it, is it really adding value?鈥�

How should young people approach AI?

David offered the finalists a memorable way to think about generative AI: Treat it like 鈥渁 friend that thinks they know everything, but they often don鈥檛.鈥�

AI systems can recognize patterns, generate possibilities and help people work more quickly, but they may not recognize gaps or biases within the information on which they rely. That makes human judgment, fact checking and verification essential.

David also encouraged students to distinguish between problems that require combining existing knowledge and those that demand genuinely new thinking. AI is particularly capable when it can blend concepts it has encountered before. Moving beyond those boundaries is much harder.

鈥淪ometimes you have to rethink the whole world from scratch,鈥� he said. 鈥淭here are a lot of ways that new ideas can form, and that鈥檚 really where a lot of the opportunity is.鈥�

Students who are comfortable pursuing 鈥渇uzzy, vague, not well-defined problems,鈥� he suggested, possess an increasingly valuable skill.

鈥淭he second the numbers are gone, most people feel lost,鈥� David says. 鈥淚t鈥檚 a rare skill to be able to work on something that鈥檚 not particularly well-defined.鈥�

Will AI replace human creativity?

David founded Midjourney in 2021 to explore the creative process itself: how people imagine new possibilities and how technology might help bring those possibilities to life.

鈥淭he most interesting question is not so much, 鈥榃hat do we want to build?鈥欌�� he said. 鈥淭he more interesting question is, 鈥榃hat do we want to become?鈥欌��

He compared working with generative AI to growing a garden. A person helps cultivate the result, but many influences contribute to what eventually emerges. 鈥淒id you make that tomato plant? I don鈥檛 know. I guess you grew that tomato plant,鈥� David said. 鈥淭hat tomato plant is a reflection of yourself in some ways, but it鈥檚 also a reflection of every tomato plant that鈥檚 ever existed.鈥�

Rather than using AI to avoid creative work, David believes people should use it to attempt more ambitious work. An artist might rapidly test ideas, experiment with color and composition, or pursue a project that would otherwise require prohibitive amounts of time.

鈥淚 think ideally we use these tools to make ourselves more ambitious and more confident,鈥� he said. 鈥淲e use these things to do bigger work that we wish we could do.鈥�

His advice extended beyond AI. Looking back, David said many of his missed opportunities resulted from judging an unfamiliar idea too quickly instead of investigating what it might offer.

鈥淟ife isn鈥檛 necessarily about giving you answers as much as it is about giving you inspiration,鈥� he said. 鈥淭he more that you can take anything that comes at you, whether it鈥檚 good or bad or right or wrong, and say, 鈥榃hat kind of inspiration for new ideas can I draw out of this?鈥� everything just works better.鈥�

Watch the .