OpenAI may be the most buzzy startup in the private markets right now, but Samir Kaul, founding partner and a managing director of Khosla Ventures, won’t say much about his firm being the first VC to back the project.
“We made a bet. And it was a large bet for us,” Kaul told me via video chat from one of Khosla’s conference rooms yesterday.
It was around 2018 when Khosla Ventures first invested in OpenAI, the Sam Altman-led startup that last month landed a multibillion-dollar investment from Microsoft. At the time Khosla invested, OpenAI was structured as a nonprofit organization, raising some questions over how exactly such an investment would have been structured. LinkedIn co-founder Reid Hoffman, who has been a partner at Greylock Partners since 2009, also invested in OpenAI at the time, though he did so personally through his foundation. (A Greylock spokeswoman says the firm doesn’t invest in nonprofits.)
Kaul discussed Sam Altman as an entrepreneur: Khosla partners are “big fans of Sam Altman,” he says, and “anything that Sam does, it automatically elevated to the top of the pile.” Kaul was also quick to point out that it was Vinod Khosla, the Kleiner Perkins veteran behind the eponymous firm he started in 2004, who led the investment, although Kaul noted the whole team worked on it and that the firm doesn’t do deal attribution.
“I don’t think about that,” Kaul said when I asked when he thought OpenAI might become profitable. He added: “What I'm more interested in is what are all the great applications of what they're going to do? How many verticals can they touch? How many different applications and products can they do? And can they continue to hire and attract and retain the best and brightest scientists? So far, that's all quite promising.”
While Kaul wouldn’t go into too many specifics, including how much Khosla has invested in OpenAI or whether the firm invested subsequently, Kaul has plenty to say about the opportunity for the technology itself and where he ultimately sees all this headed. He believes generative A.I. will ultimately influence every sector—from enterprise to consumer to financial services. And, of course, to health care—where Kaul has spent much of his career. (Kaul has started several companies, and he used to work for Craig Venter, who is known for pioneering the use of automated gene sequencing. Kaul helped sequence the first plant genome to ever be sequenced: a mustard seed called Arabidopsis thaliana.)
Health care is where Kaul really lights up when talking about the future of A.I.
“So when you go to the doctor, you probably notice your physician has your folder that he or she has probably looked at for maybe 13 seconds before you come in, and they'll ask you a bunch of questions while they're reading through the form,” Kaul says. They haven’t prepped, he goes on. They are seeing, just now, that you had a vaccine last year, but not this year. You had this surgery, this illness, and here is your family history, and here are the meds you take. “They're flipping through it right there because they give you about six minutes, right? You're in and out, because they know that's expensive. They have to pay tuition. They have to pay loans. I got to get malpractice insurance to pay the overhead of wherever the hospital is or whatever practice they're at. It's not easy.”
Physicians can’t digest all that data in seconds, he argues, and they can’t digest all the research that's being published on a weekly basis in terms of diagnostics and medicine.
“A.I. can ingest all of that data in seconds,” Kaul says. There’s so much buzz around search or replacing Wikipedia or robotics for mechanizing labor, “But, just to me, what's really exciting is to have an A.I. doctor that knows my entire system, has all 49 years of my data, every med I've ever taken, every surgery I've ever had, all my genome sequencing—and that is constantly blasting that against all the new publications and all new research and all new clinical trials and giving me recommendations.”
Kaul pointed me back to a paper that Vinod Khosla authored in 2016—which made a rather bold claim that 80% of every physician’s workload may end up being replaced by technology as the health care world entertains more “inexpensive data-gathering techniques, continual monitoring, more rigorous science and more available and ubiquitous information leading to personalized, precise, and consistent (across doctors) insights into a patient.”
And if that’s the kind of technology a company could help develop, Kaul says the firm is happy to wait to see a return. Referring to all Khosla’s big swing investments, which have included bets in fintechs like Square, Affirm, or Stripe; consumer companies like Instacart and DoorDash; enterprise companies like RingCentral or GitLab; and health care companies like SWORD Health, Hello Heart, and Ginger; or food companies like Impossible Foods, Kaul says: “We will be patient. We don't mind keeping a company at, you know, a $200k a month burn for four years to get it to the right point.”
Given the rather unusual structure of Microsoft’s most recent investment in OpenAI, Khosla might end up having to wait a while.
The big Bitcoin whale…It was about six years ago that a professor at the University of Texas McCombs School of Business sifted through 200 gigabytes of trading data to determine how an unidentified Bitcoin “whale” distorted trading and almost singlehandedly drove the token’s run-up between late 2017 and early 2018. Is it happening again now? Find out in Shawn Tully’s new feature here.
See you tomorrow,
Jessica Mathews
Twitter: @jessicakmathews
Email: jessica.mathews@fortune.com
Submit a deal for the Term Sheet newsletter here.
Jackson Fordyce curated the deals section of today’s newsletter.
