Khosla explains his 'robots replacing doctors' comment and goes on the hunt for data scientists

REDWOOD CITY: Vinod Khosla’s first bone to pick is with technology journalists, whom he refers to as “English grads who don’t understand English.”

His second is with modern medicine, which he believes is more of a practice rather than a science.

Khosla is a rare Silicon Valley type who does not hold back — in his investments, blog posts, or interviews with the press. He hit the headlines last year (and sparked indignation in the medical community) when he reportedly said at a conference that 80 percent of doctors would be replaced by technology. At VentureBeat’s DataBeat/Data Science Summit, Khosla stressed that he stands by this comment, which was widely misunderstood by “English majors,” his term for reporters he does not like.

Khosla does not think we should wipe out medical practitioners in favor of smart machines. He believes that medicine is complex, and human beings aren’t designed to tackle such complexity. In the future, new technologies will take a lot of the guesswork out of medicine and assist doctors in various functions, including surgery, diagnosis and patient monitoring.

“By 2025, 80 percent of the functions doctors do will be done much better and much more cheaply by machines and machine learned algorithms,” he said. He also stressed that this is a probabilistic statement, so it is merely something he suspects to be true.

Khosla throws out a few harrowing statistics to illustrate his point: Most of what medical researchers conclude in their studies is misleading or exaggerated, the Atlantic reports. “Most of these studies are biased,” said Khosla. “They want to get through the FDA’s [the Food and Drug Administration] process — not discover the real facts.”

Khosla’s firm, aptly named Khosla Ventures, is one of the most active firms investing in new health care technologies and tools. He believes that the first wave of this new health care technology is services like Practice Fusion and ZocDoc (a KV portfolio company), which work with care providers to migrate paper-based processes online. The next generation is robotics, sensor networks, and complex algorithmic tools to make sense of a mountain of health care data.

“Your cellphone has 10 sensors, and your car has 400. But your body has none — that’s going to change,” said Khosla.

A sensor might track an individual’s basic body metrics while they sleep. In the event of a blood pressure spike or other abnormal event, that person might receive a notification to see their doctor. Invetors will need to come up with machines to manage a vast number of data points about each person per day and turn this into useful information and alerts. “We will have millions of times more data about patients,” said Khosla. “We could use this data to prevent disease.”

Khosla predicts that these new data services will also flag drugs that are doing more harm than good, once approved. For instance, beta blockers were hailed as a life-saving drug, but we later discovered that the side effects were serious, even deadly. By tracking patients more closely through sensor devices, we could track how each person’s body is reacting to a drug, and pinpoint the dangerous ones.

With this thesis in mind, Khosla is on the hunt to hire data scientists who are interested in health care.

“Data scientists: Send me your resumes. Entrepreneurs: Send me your business plans. VCs — I won’t talk to you,” he joked.