Gordon Hirsch Wilson and the art of making a difficult idea travel

Gordon Hirsch Wilson’s account of building an AI chip company includes a useful complication: the company had to change its mind. Rain began with an analog vision, pursued it through two chip tapeouts, and concluded that the necessary materials were not ready. The ambition survived. The route to a product changed. For a founder who describes himself as a storyteller, this was a demanding editorial assignment. The plot had acquired evidence.

Wilson helped found Rain in 2017 with Jack Kendall and Juan Claudio Nino, whom he met at the University of Florida. He led it for six years, then stepped down as CEO in 2023 and announced a move into an executive advisory role. His subsequent work has included an online bookstore and an AI investment firm. Across those settings, his own description of the job has been unusually consistent: help complicated ideas make sense to other people.

First, persuade a voter

Before chips came campaigns. Wilson worked in political campaigning and nonprofit advocacy in Central Florida. His experience included a congressional race, management of a state legislature campaign, and work at Equality Florida. In his account of that period, he managed the largest field office of a contested congressional election in 2010 and helped the team that elected Florida’s first openly LGBT state representative in 2012.

Those are jobs with a fairly unforgiving measure of communication. Someone must understand a message, care about it, and eventually do something. A sentence that impresses the person who wrote it has accomplished very little on its own. It is a useful earlier chapter for someone whose later work would involve explaining why a computer might need a different physical architecture. The audiences changed; the need for a reason to pay attention remained.

At the University of Florida, Wilson studied mathematics and statistics. He and Kendall became successive presidents of a student organization teaching Python and machine learning. Kendall’s interests crossed physics, chemical engineering, materials research, neuroscience, and AI. The student group gave them a place to meet across those boundaries. Rain grew out of that university setting, rather than an established semiconductor company’s product department.

A bookshelf before a business plan

Wilson grew up around entrepreneurs and science fiction. Both parents were entrepreneurs; the family’s bookshelves made room for imagined futures as well as everyday business. In a later interview, he recalled books ranging from Rendezvous with Rama to Hyperion. Building an organization of his own had seemed an appealing possibility long before there was a Rain to build. The novels supplied questions about the future. A company would require rather more paperwork.

The intellectual attraction at Rain was a machine that could learn efficiently. Wilson and his co-founders wanted to take clues from biological intelligence and use them to rethink computing hardware. That meant bringing people with very different specialisms into the same project. The scientific question was large. So was the practical question of how those people would work together without spending their lives explaining their vocabulary.

“My failures are just as essential to my growth as my successes are.”

Gordon Hirsch Wilson

The distance between memory and thought

Rain entered Y Combinator’s Summer 2018 batch. Its early approach used memristive nanowire networks: physical structures intended to perform neural-network operations, rather than simply represent them in software. The ambition was a device with a learning system built into its hardware. This was an argument about the machinery underneath AI, at a time when much of the public conversation was about what software could do on top of it.

Wilson repeatedly returned to the expense of moving information. Conventional computing separates memory from processing. Rain’s compute-in-memory approach sought to bring those activities together. The attraction is easy to picture: a worker who can perform a task at the filing cabinet spends less time fetching folders. In a chip, the folders are data. The journey has an energy cost, repeated across enormous numbers of operations.

His 2020 conversations about the technology covered analog computation, nanowires, programmable resistance, and the challenge of scaling neural networks. These are connected problems, but they belong to different technical conversations. A materials decision can affect a circuit; a circuit can constrain an algorithm. For Wilson, explaining the larger system was part of making those individual decisions intelligible.

An interpreter with a fundraising deadline

Wilson’s description of his CEO role included two occupations: storyteller and translator. He needed to connect device physics, circuit and system architecture, algorithms, and software. He also needed to translate technical work into the value it might offer someone outside the laboratory. A good explanation had to survive both audiences. Investors needed a reason to fund it. Engineers needed an account that still resembled the engineering.

That role helps explain why his public talks spend time on the path from the brain to the processor. The history gives an audience a way into a proposal that otherwise arrives surrounded by specialist terms. Wilson’s 2021 conference biography made translation central to the work. It also placed his life in San Francisco, with interests in swimming, weightlifting, and his local community. A person can worry about semiconductor architecture and still have plans after work.

In February 2022, Rain announced a $25 million Series A led by Prosperity7 Ventures. The announcement named existing backers including Buckley Ventures, Gaingels, Loup Ventures, Metaplanet, and Pioneer Fund. Sam Altman had led the company’s 2018 seed round. Rain also said it had taped out a working prototype in 2021. These were concrete steps in a hardware company’s development, with more engineering and commercial work still ahead.

Wilson also appeared on the steering committee for the 2022 International Conference on Neuromorphic Systems. The listing puts a company founder alongside researchers at national laboratories and universities. It is another small indication of the territory Rain occupied: commercial development with one foot still in research. Turning an idea into a product did not remove the need to keep asking scientific questions.

In a 2019 conversation, his recommended reading included Jeff Hawkins’s On Intelligence and Tim Urban’s Wait But Why. The pair make a fitting reading list for someone interested in both intelligence and explanation. One explores how minds work; the other has made a practice of giving complicated questions enough space to become readable.

When the materials refused the script

Wilson later gave a candid account of what happened to the original analog plan. Two tapeouts had exposed a gap between the concept and the maturity of the materials needed to realize it. Rain redirected its product roadmap toward digital SRAM-based compute-in-memory. Advisors, including Georgia Tech’s Arijit Raychowdhury, helped shape that reassessment. The company retained research into more experimental materials, architectures, and algorithms.

That decision preserved the interest in bringing memory and processing closer together while changing the substrate used to pursue it. There is a revealing distinction here between a founder’s commitment to a problem and commitment to a particular answer. Wilson’s account acknowledged that the original answer needed more time. Semiconductor materials have a certain immunity to a persuasive presentation. They remain an exacting audience.

By late 2023, he had also changed his own position. In announcing his departure from the CEO role, Wilson recalled that he and Kendall were 25 when they started, with no prior experience building companies or bringing novel semiconductors to market. He described a team approaching 40 people and thanked the specialists across chip design, software, physics, and AI who had joined it. He would continue helping as an executive advisor.

The next chapter sold books

Wilson’s move into Allstora, announced in 2024, brought him into a different sort of storytelling business. He said he had joined the online bookstore as chief strategy officer that January. Allstora’s founders included historian and author Eric Cervini and RuPaul Charles. Wilson described the work as supporting authors and underrepresented voices, with AI used to help writers earn and create. His stated interest was in tools that served the people making the stories.

The move also connected technology with an older interest in advocacy. He highlighted the Rainbow Book Bus, Allstora’s philanthropic arm, and its distribution of diverse literature to communities facing book bans. It was a change of product, colleagues, and industry. Yet the people at the center of his announcement were writers trying to reach readers. For someone who had spent years trying to make difficult ideas accessible, the appeal is understandable.

Still looking for people to talk to

HalfCourt now lists Wilson as a venture partner, identifying him through his experience as Rain’s founding CEO and his University of Florida background. The firm focuses on AI. His place on that team brings the experience of starting and leading a hardware company into conversations with other founders. It is another setting in which an unfamiliar proposition needs to be understood before anyone can sensibly decide what to do with it.

There are lighter footnotes. Wilson has said he acted in a five-minute film about Grindr that screened at Cannes. Dogs reliably win his affection. In a public travel announcement, he invited conversations around the world about how different cultures experience AI: what people welcome, what they fear, and what they expect. That is a rather more interesting itinerary than a succession of conference lanyards.

His last presentation as Rain’s CEO took place at Ai4 on August 8, 2023. When he later shared the recording, he described it as an example of his approach to accessible technological storytelling and distinguished his own perspective from Rain’s current positions. The distinction suits the broader career. Ideas acquire new evidence. Companies change their plans. A useful storyteller keeps listening, and leaves enough room in the account for the next thing that turns out to be true.