Inspired with Alexa von Tobel
00:02:27 · Alexa von Tobel, Mikey Shulman
[00:02:27] S1: Awesome. So Mikey, I wanna start from the beginning just really quickly between Kensho and now Suno. I wanna just start from your childhood. You grew up in New York. Is there anything that stands out? I always like to ask the question just sitting here as a parent with three kids of, is there anything that stands out in your childhood that you feel like has set you up for kind of where you are today? [00:02:49] S0: You know, I think about this a lot. I think I had just such an incredibly lucky upbringing. And, honestly, you know, the more I do it, just realizing that coming from a a loving, stable environment is, the best leg up you can have. And, you know, I think if you force me to pick one thing, what was modeled for me is, like, going after what you love to do. And maybe that changes as you grow up, but I I, you know, I can tell you, like, probably the earliest tangible or important one that I remember is just... I remember when I was... When I realized that I really liked physics and just having my parents encourage me, like, go do that. That's cool. If you like it, just go do it. Doesn't doesn't matter what's gonna come after, just go do it. And I think that's something that I've taken with me, like, through everything. I don't know if there's ever, like, a big master plan. It's just you like it, keep go doing it, you know, just go tackle the next thing that's in front of you.
00:02:49 · Mikey Shulman, Alexa von Tobel
[00:02:49] S0: You know, I think about this a lot. I think I had just such an incredibly lucky upbringing. And, honestly, you know, the more I do it, just realizing that coming from a a loving, stable environment is, the best leg up you can have. And, you know, I think if you force me to pick one thing, what was modeled for me is, like, going after what you love to do. And maybe that changes as you grow up, but I I, you know, I can tell you, like, probably the earliest tangible or important one that I remember is just... I remember when I was... When I realized that I really liked physics and just having my parents encourage me, like, go do that. That's cool. If you like it, just go do it. Doesn't doesn't matter what's gonna come after, just go do it. And I think that's something that I've taken with me, like, through everything. I don't know if there's ever, like, a big master plan. It's just you like it, keep go doing it, you know, just go tackle the next thing that's in front of you. [00:03:54] S1: A loving, stable environment is a real gift. That's beautifully said. Mikey, you and I overlapped briefly at Harvard. I wish I knew you a lot better when you were there. We would have had fun together. But you earned your physics degree at Columbia, then you completed a six year PhD at Harvard, which was 2009 through 2015, focused on experimental quantum computing. So you're a real dumb dumb. How did you decide to go down the PhD path? Just tell us a little bit about how you sort of thought about your academic career.
00:04:26 · Mikey Shulman, Alexa von Tobel
[00:04:26] S0: It's really it's really interesting. I always loved physics. I knew I loved it in in high school even. And I was trying to decide as an undergrad what I should do after. And it's like one of these, like, seems little at the time, but very serendipitous moments where there was a professor that I wanted to work for the summer after my junior year of college, and he said, I'm retiring. You can't really work for me. I'm gonna be closing down my lab. And then I went and I applied for an internship at Goldman Sachs. And then two months later, he said, you know, I'm actually gonna do one more year. Why don't you come work for me? And that was, like, this little thing that sent me down an entirely different career path where I was like, I really loved working for him, and I really loved the work, and it made me wanna do that. And then after that summer, said, okay. I'm gonna apply to grad school. Like, I I wanna do more of that. And I I suspect that if that hadn't happened, if I hadn't worked for him that summer, it would have been a very different path for me. And, yeah, you know, the rest is history. Actually, my PhD adviser is a former student of that undergrad advisor, and he told me just go work for your PhD advisor. He's the best student I ever had. [00:05:41] S1: That's amazing. You then went from... Instead of staying in academia, you jumped right into Kensho where you led machine learning for financial systems. And Kensho was really your first major step into applied AI. How'd you... How how how'd that happen? And also tell us what Kensho was.
00:05:41 · Alexa von Tobel, Mikey Shulman
[00:05:41] S1: That's amazing. You then went from... Instead of staying in academia, you jumped right into Kensho where you led machine learning for financial systems. And Kensho was really your first major step into applied AI. How'd you... How how how'd that happen? And also tell us what Kensho was. [00:06:00] S0: Yeah. You know, it's funny. So Kensho is and was an AI company in financial services. And I think very much the same thing. I didn't have a master plan at the end of my PhD. I met the Kensho people. The team was probably 12 when I joined, and I met all 12, and I liked all 12. And I said, I like working with you guys. Will you make me a job offer? Or I interviewed, and I actually didn't do very well in the interview, but they were kind enough to make me a job offer. And, you know, it's funny in a in a in a small company, you know, with 12 people, there's, like, no access controls on the on the applicant tracking system. And so, actually, after I joined, somebody just showed me what everyone wrote about me, and it wasn't it wasn't terribly kind. I actually did not do very well in those interviews, and it was very nice that they that they took a they took a flyer on me. [00:06:52] S1: You're an easy guy to take a flyer on, by the way.
00:06:52 · Alexa von Tobel, Mikey Shulman
[00:06:52] S1: You're an easy guy to take a flyer on, by the way. [00:06:55] S0: I'm not. I'm... You know, that's very kind of you to say. I I think, like, I... Yeah. And then I was hired as a software engineer, and very quickly, the the machine learning opportunities came along to the company. And this is back in 2014 and 2015 when AI was, like, a much less developed field. And I guess the the guy with the PhD in physics is what passes for machine learning engineer back then. And so I naturally gravitated toward those things. And I think I very quickly realized, a, I like it, but b, actually, where my... I think where my real talents lay were less in the day to day work, which I was, like, okay at, but more in building the team and identifying people who are excellent at it. And so I was honestly given a very long leash by Daniel, the founder, to go and build that team as I wanted and took full advantage of that, and kind of the rest is history. And, you know, I enjoyed every minute of it. [00:07:51] S1: Tell us a little bit about the outcome of Kensho and what happened, because it was a really successful outcome for you guys.
00:07:51 · Alexa von Tobel, Mikey Shulman
[00:07:51] S1: Tell us a little bit about the outcome of Kensho and what happened, because it was a really successful outcome for you guys. [00:07:57] S0: Yeah. It was... Look. It was it was a lot of fun, and I will say we're... I'm currently calling in from our headquarters in Cambridge, Massachusetts, which is the same office that Kensho's first office was, and that... That's somewhat by design. We were acquired in 2018 by S and P Global, a huge financial services company, after we were working together with S and P on various projects and realizing that there was a ton of really important work to do together around, at a very high level, bringing a lot more structure to the crazy world of unstructured documents in financial services. And so, you know, maybe just as an example, you read Nvidia just beat consensus for earnings, and, you know, maybe you never think what does that really mean. What does that consensus really mean? That's like people going and reading all of the broker research from 10 different banks and pulling out numbers from hundreds of pages of documents and averaging those numbers, and how messy of a process that is, and how much more structure can be brought to bear if you use the right set of set of tools. And so we kind of developed those with S and P first at arm's length, and then after we were acquired, not at arm's length. And I think ultimately ended up bringing a lot of speed and scale to that business. And so I think a successful outcome for both sides, which I'm told is, like, somewhat rare in the game of acquisitions.