Ren Ito LLM ’04, co-founder and COO of Sakana AI, explains how his legal degree has helped him in a non-legal career

Ren Ito LLM ’04 expected to practice law after completing his LLM, but other opportunities kept getting in the way. First, it was the chance to work in the political office of the Japanese Embassy in Washington, DC. Then it was the opportunity to team up with an old high school classmate to run the e-commerce company Mercari—Japan’s first unicorn, or privately-owned start-up valued over $1 billion dollars. As the CEO of Mercari Europe, Ito facilitated its $6 billion initial public offering in 2018. In 2022, major developments in artificial intelligence convinced Ito to take on the role of chief operating officer of Stability AI, a UK-based unicorn startup, which he held until last summer. Now, after launching his own generative AI company, Sakana AI, in January, Ito is finally ready to admit that traditional legal work may not be in the cards.

Ren Ito
Ren Ito

“At every point, I intended to begin practicing law, but I might be a bit too old to start as a first-year associate now,” Ito quips. “But in many ways, I use my law school degree in all my ventures. My legal education trained me in consistency, reciprocity, and proportionality, all of which I draw on constantly,” he says.

In this Q&A, Ito talks about the versatility of a law degree and how pursuing emerging trends has led to a fulfilling career.

How did you decide on NYU Law for your LLM degree?

When comparing different law schools, there are often unique attributes that make each stand out. For NYU, that quality was its pride in being a global law school. This is a principle I really resonated with. As a diplomat in Japan, I developed an interest in international arbitration and maritime issues [and] so, I wanted to pursue an LLM degree that was focused internationally.

Once at law school, in addition to my international interests, I became really interested in American domestic law. This interest in American civil procedure and constitutional law led me to take a job at the embassy in DC after graduation.

What did you like about working as a diplomat? How did you transition from foreign service to the technology space?

What I liked most was the almost-journalistic nature of the foreign service; to be a good diplomat, you had to excel both at making new friends and connections, and at collecting and analyzing information. Of course, there were all the exciting opportunities that came with the job too: at one of President Obama’s meetings with the Japanese Prime Minister, I was the only other person present; and I shook hands with him, many times! During this time, my high school buddy started Mercari and then asked me to join. I was with him for the entire journey, from the start to the IPO. It was such a great experience that I decided to stick with the tech industry afterwards as both an operator and investor. I ended up being involved with four different unicorns, two of which were in the artificial intelligence space where I am now.

For me, one of the consistencies with being in the foreign service, a law student, or in the tech space is a love of troubleshooting. Troubleshooting is the essence of what you do in law school, and that is definitely true in my line of business. I was never a general counsel in the start-up space, but my overall approach based on legal expertise proved incredibly useful.

What do you see as the potential benefits of generative AI?

Industries will be—and if anything, already are—the first to reap the benefits of generative AI; as we are seeing rapid adoption and rising use cases across several sectors, generative AI is set to change the very nature of work. This technology undoubtedly has the potential to increase productivity depending on how it is used, but there are also major questions to still be addressed about how people will need to be skilled differently in this generative AI age.

To feel the real benefit that will arise from this wave of generative AI, however, we need to wait for the rise of killer apps. The example that I find useful to think about this is Windows 95. When Windows 95 was coming out, we had no PowerPoint and no Excel. We had the operating system, but we had to wait for the killer applicationss that directly address specific use cases to make that new system more meaningful. I think we are in the same place with generative AI. We are waiting for the next generation of apps and I am sure we won’t have to wait too long.

Currently, artificial intelligence is becoming mainstream, and it’s moving fast. Technology is evolving at a pace that really outstrips the government’s or court’s ability to understand it, so it is increasingly the case that the industry and startups and tech companies are setting their standards of use. You watch unwritten laws begin to develop, and you have to be really agile in setting the rules and the standard. That said, the answer is not necessarily to cut governments out of the equation. On the contrary, we are seeing increasing value placed on public-private partnerships, to ensure a dynamic approach to regulation that both protects the public in their use of generative AI and inspires groundbreaking innovation.

Will you stick with the tech space? Or do you see something new on the horizon?

I would love to do so. There have been no winners identified in the AI space yet, and so it feels like an exciting time.

This is partially why I recently co-founded Sakana AI. At Sakana, we believe that there is an alternative path to innovation in generative AI, one that can deliver equally powerful models as the current leaders in this space, but at a [much lower] cost.

We achieve this through a philosophy called nature-inspired intelligence. Perhaps the easiest way to explain this is through imagining a bee colony. The worker bees leave the hive to forage for resources, and they can then communicate with each other about where the good resources are located, thereby making it easier for the other bees and benefiting the colony overall. This is exactly how Sakana’s model will operate: through drawing knowledge from the outputs of existing models, Sakana will collate and synthesize the best elements of various responses to a given prompt. We hope to provide what may be seen as an ideal form of collective AI-generated information.

I am also getting more and more interested in legal education. Right now, I’m actually working as a senior fellow with the Law School’s US-Asia Law Institute. Though I am neither a US citizen nor have I practiced law directly, I have learned and benefited a lot from law school and want to help bring that kind of diversity to NYU Law. Legal training is extremely helpful, and I believe it also helped formulate my sense of integrity in my work.

What have you found particularly rewarding or challenging about your work thus far?

My career has been the product of strange turns of events; I never could have predicted that I would be working in AI now, let alone the rest of my career trajectory. But if you follow what interests you and what is emerging, I believe you will not struggle to find enjoyment and excitement in what you do.

This interview has been edited and condensed. Posted March 19, 2024.