Why you should care
Because she’s got brains and access to robots.
Shivon Zilis is the embodiment of sprezzatura, wearing little to no makeup, jeans and a tee, even in the chichi enclave of San Francisco’s private club the Battery. She keeps her quilted red jacket on as we talk, and I’m occasionally distracted by the text on her graphic shirt. All I can see is the word sex. I finally get up the nerve to ask what it says. Good grammar is sexy.
Ah, San Francisco, where a 29-year-old can be one of the tech world’s most promising investors and still dress like the cool tomboy next door. Zilis is one of the founding members of 2 1/2-year-old Bloomberg Beta, an early-stage venture capital fund whose investor is Bloomberg LP, and she specializes in what industry watchers consider one of the most important subfields that’ll power the next wave of the boom: machine learning. That means data and stuff, sure, but also artificial intelligence (listening yet?), self-driving cars and even IBM’s Watson, the computer that killed it on Jeopardy!
Those are the easily digestible things that come from machine learning. Behind the scenes lies a nascent world with some 2,500 startups jockeying for space, according to Zilis’s count. And sitting at the helm of a $75 million fund, Zilis has the chance to pick her potential winners. That’s a lot of power for a millennial. Especially given that this is an industry about to blow up; that favorite jargony term big data is forecast to be a $24 billion industry next year, and much of that data will come from machine learning. This stuff is probably “the most important thing in our overall investing plan,” says James Cham, a Bloomberg Beta partner.
Zilis rarely pauses for breath, speaking in fully formed paragraphs sans flourish. Her eyes are big, icy and piercing, to start with. As she picks up speed, they grow even wider. (Born to an Indian-immigrant mother and a Caucasian father in Canada, Zilis “pretty much turned out white — I just got the big eyes from the Punjabi side,” she says.) A former college hockey player who in another lifetime would have spent her career moneyballing the sport, she’s no-bullshit, calling herself the “unsexy investor” who opts for agricultural technology over Snapchat. She does come from a down-to-earth professional background: a job at IBM after college (she was delighted to see Watson’s success!) and then to Bloomberg, where she built internal startups and worked on partnerships between Bloomberg and the startup world.
Which gets to her praxis. Talk to most investors about how they do what they do, and you can prepare yourself for an onslaught of mild self-congratulation and vagueness. Venture capitalists are like mythological cowboys out here, bragging of instinct and some magical ability to just intuit a founder’s ability. Zilis, on the other hand, says she’s “thesis-driven,” a cerebral evaluator, “not a networker.” Of course, it’s not a dichotomy between the people-people and the nitty-gritty people, but on a spectrum, Zilis falls toward the latter end.
She identifies with Hume and his “high-level philosophy, which also touches earth.”
That bird’s-eye view hasn’t always worked out for her, she admits. She says that when you focus so much on the big ideas and geek out on a field, you might end up knowing more than a founder who’s pitching you — and she has projected her own guesses at what a company should be doing onto a founder who wasn’t quite so adroit in the field. What the skill has done is given her the chance to sort of reverse-pitch founders who have their elevator pitch down but “don’t have the two-minute version,” Cham says. Newsle co-founder Jonah Varon, whose company was acquired by LinkedIn (one of Zilis’s most successful exits), says Zilis is not quite as nerdy as she claims to be — “a unique combination of IQ and EQ.”
Bloomberg Beta backs early-stage companies, but in the machine-learning world, some stuff is too fetal for even Zilis. Much of the science is still in research labs, way pre-business model. Artificial intelligence is “almost ready for business” but “very emerging,” says industry watcher Brad Power, a consultant and researcher on process innovation. Depending on how you look at it, it’s either a tough time to be parsing it for opportunity or the best time ever; Power figures it’ll blow up in the next year.
Apparently, the intuition-plus-precision model is how Zilis’s brain has always worked. While indulging in varsity athletics (Zilis, absurdly, also tried to join the fencing team at the same time as hockey), she majored in economics and philosophy at Yale. Her two favorite philosophers: Nietzsche and Hume. Nietzsche: flowery, lovely. She identifies more with Hume and his “high-level philosophy, which also touches earth.”
She recalls always having some foresight: A friend would come back from a job interview or talk about a new boyfriend, and she “could just tell” if she’d gotten the gig or if the relationship would last. Sounds like tea-leaf reading stuff to me, but she’s far too rational for me to simply accept it as uncanny. Philosophy, she says, gave her a logical backbone. But it still has a whiff of the mystical about it. “I always knew that stuff was going to happen before it happened.”