Vinod Khosla: Technology Will Replace 80 Percent of Docs

Vinod Khosla: Technology Will Replace 80 Percent of Docs


I recently viewed health care through the lenses of a technology entrepreneur by attending the Health Innovation Summit hosted by Rock Health in San Francisco. As a practicing primary care doctor, I was inspired to hear from Andy Grove, former CEO of Intel, listen to Thomas Goetz, executive editor of Wired magazine, and Dr. Tom Lee, founder of One Medical Group as well as ePocrates.

Not surprising, the most fascinating person, was the keynote speaker, Vinod Khosla, co-founder of Sun Microsystems as well as a partner in a couple venture capital firms.

“Health care is like witchcraft and just based on tradition.”

Entrepreneurs need to develop technology that would stop doctors from practicing like “voodoo doctors” and be more like scientists.

Health care must be more data driven and about wellness, not sick care.

Eighty percent of doctors could be replaced by machines.

Khosla assured the audience that being part of the health care system was a burden and disadvantage.  To disrupt health care, entrepreneurs do not need to be part of the system or status quo. He cited the example of CEO Jack Dorsey of Square (a wireless payment system allowing anyone to accept credit cards rather than setup a more costly corporate account with Visa / MasterCard) who reflected in a Wired magazine article that the ability to disrupt the electronic payment system which had stymied others for years was because of the 250 employees at Square, only 5 ever worked in that industry.

Khosla believed that patients would be better off getting diagnosed by a machine than by doctors. Creating such a system was a simple problem to solve. Google’s development of a driverless smart car was “two orders of magnitude more complex” than providing the right diagnosis. A good machine learning system not only would be cheaper, more accurate and objective, but also effectively replace 80 percent of doctors simply by being better than the average doctor. To do so, the level of machine expertise would need to be in the 80th percentile of doctors’ expertise.

Is it possible technology entrepreneurs can disrupt health care? He challenged any doctor in the room to counter his points.


Was it because everyone agreed? Were the doctors in the room simply stunned? Was there a doctor in the house? And where did he get that 80 percent statistic?

Was Kholsa serious that technology could make health care better by utilizing large data sets and computational power to clinch better and more precise diagnoses?  Was he simply being provocative to hear other points of view to learn even more? Like many others in the conference, he believes that giving consumers more opportunities, access, and choice to information about themselves and their bodies would empower them to do the right thing. He held up an EKG attachment to the iPhone which was just one of many consumer directed products in the pipeline his company has invested in.

Kholsa is a very smart and successful entrepreneur. Does innovation mean the two guys in a garage who come up with a radical idea or is it possible that innovation is having people with different experiences and point of view looking at the same problem as best selling author Malcolm Gladwell noted in his New Yorker piece Creation Myth –  Xerox PARC, Apple, and the Truth About Innovation? Surely to make health care better, technology entrepreneurs must engage with doctors. All the speakers before and after Kholsa spoke about the incredible value and insight different stakeholders to bring to the table.  The most vocal? The doctor entrepreneurs and those who worked with doctors to bring their ideas to market.

Kholsa’s criticism of the health care system is completely valid. Can we do better in being more reliable, consistent, and creating a system process and design that is comparable to highly reliable organizations and industries? Of course. Can we be more systematic and doing the right things every patient every time on areas where the science is known to level of the molecule? Yes. Care must be incredibly simple to access, extremely convenient and intensely personal.

It isn’t that we don’t have smart people. Compared to a century ago more illness are understood, specific medications and treatment protocols can be designed. But we haven’t solved it all. When we thought we knew it all, we were shown how little we truly knew. Thirty years ago, doctors predicted the demise of infectious diseases as a specialty, another footnote in medical texts as more powerful antibiotics and vaccines were available. Enter AIDS, the swine flu, and many super-bugs which have humbled our profession. Ask accomplished physician and writer Dr. Abraham Verghese about his experiences. Dr. Verghese is rightly worried, as many others are, that even doctors are being too focused on the iPatient and not on the real patient as he writes in his New York Times op-ed Treat the Patient, Not the CT Scan. Is this what we want our health care system to look like?

Health and medical care is an incredible intersection of technology, science, emotions, and human imperfections in both providing care and comfort. As conference speaker Dr. Aenor Sawyer, an orthopedic surgeon from UCSF noted, we need to figure out how to have our different cultures of doctors, gamers, designers, and technologists interact. Fixing health care is more than simply “we know the problem and we know the solution”. She reflected that the level of dedication, perseverance, and a willingness to make impact among the different groups demonstrates more similarities than differences.

I know health care can’t simply be solved by smart people in Silicon Valley alone.  To solve health care we need everyone to collaborate. As Harvard Business School professor Amy Edmondson noted in her book Teaming

“For over a century, we’ve focused too much on relentless execution and depended too much on fear to get things done. That era is over…human and organizational obstacles to teaming and learning can be overcome…Few of today’s most pressing social problems can be solved within the four walls of any organization, no matter how enlightened or extraordinary… Generating ideas to solve problems is the currency of the future; teaming is the way to develop, implement, and improve those ideas.”

Perhaps Kholsa’s call to action was simply an entrepreneurial mindset, but simply ignoring those who have chosen a field to improve and safe lives and who meet humanity everyday on the front-lines is problematic and dangerous. There are some things that may never be codified or driven into algorthims. Call it a doctor’s experience, intuition, and therapeutic touch and listening. If start-ups can clear the obstacles and restore the timeless doctor-patient relationship and human connection, then perhaps the future of health care is bright after all.

Davis Liu, MD, is a practicing board-certified family physician and author of the book, “Stay Healthy, Live Longer, Spend Wisely – Making Intelligent Choices in America’s Healthcare System.” Follow him at his blog, Saving Money and Surviving the Healthcare Crisis or on Twitter, davisliumd.

Leave a Reply

90 Comments on "Vinod Khosla: Technology Will Replace 80 Percent of Docs"

Mar 27, 2015

Some procedures may be replaced by machines, but doctors will not. In my opinion of course.

Jan 14, 2015

Nice article has ideas, but the body is 1,000,000 times more complex than putting a man on the moon. How much did it cost per trip? My point is a new network came to town and wants to pay $22 dollars a visit. We have saved many lives last year that if the patient had not changed to us the patients would be dead today.
This was done by listening to the patient performing tests that were not previously ordered because they were cost cost continence. These patients lives were saved as a result of the Dr spending more time with the patient. Many times an hour.
Please explain to me how I pay $350,000 in school loans, 20,000 in Malpractice insurance, Rent, Staff, and my living expenses. Medicaid pays $68 for an hour visit. Traditional insurance pays $114. Yes I could see 15 patients an hour and not save a one of them. If your loved one gets sick and you have two options. Wait several weeks to get your loved one an appointment, and wait in the lobby for up to four hours and just pay your $25 co pay, See the doctor for 5 minutes or Pay $200 to get in today, Wait 5 minutes in the lobby. I know what I am going to do.
Ah you are thinking Nurse Practitioner. How many times do you take your Porsche, Benz to Walmart for a oil change. A doctor goes to school for many many years. A nurse practitioner knows perhaps at best 16% of what a doctor does. My best estimate. is many patients will die unless you spend time with a good doctor.
I have a fix that no one address email me at


When Should electrical contractors in san diego a GFCI Outlet be Installed?
A certified electrician from well known companies would send to customers’ homes would have a fire or
other safety hazards. You need to also check out all electrical contractors in san diego your payment terms.

Hiring a company to do electrician labour in their own electricians.
Compensation is thus guaranteed in this case as an electrician’s apprentice.


I think this is one of the most important information for me.

And i’m glad reading your article. But want to remark on some general things,
The web site style is great, the articles is really great :
D. Good job, cheers


Hi there, just turned into alert to your weblog through Google, and found that it is truly informative. I’m going to be careful for brussels. I’ll be grateful for those who proceed this in future. Numerous other folks will probably be benefited from your writing. Cheers!


My designer is wanting to be able to persuade myself to go to be able to .net coming from Perl. We have always resented the thought due to the bills. Nevertheless he is tryiong none the much less. I have been making use of Hubpages on the selection of sites approximately 12 months and ‘m focused on changing to another program.

Dec 5, 2012

Just coming back to this as I heard Vinod Khosia speak today. I think Carolyn brings up a point that many in healthcare (physicians and others as well) ignore – diagnosing isn’t perfect, and when healthcare professionals are criticized, they often don’t learn from it but instead get defensive. I don’t know the answers to this, but surely we must agree that many consumers are not happy with the current system, whether for cost or outcomes reasons, and we need to be open to making things better. If that takes technology, great. If it takes a different payment system, great. But let’s not pretend it’s all perfect.

Dec 11, 2012

Sandra has highlighted vital point. This is particularly the case in case of Diabetes.
We have developed algorithm based program for Diabetes Complications Management. It is Patient-centric and goes through the process of Diagnosis – Prognosis – Guidance (Medical Nutrition Therapy / Therapeutic Lifestyle Modification / Exercise / Medication check-list) which would empower the Physician as a reference point for his Medical prescription.
The system works through thousands of permutations and combination, which humanly is impossible to contemplate. The program incorporates CDSS features.
This is one of the ideal example of how Technology complements Physician skill by ensuring that all the established protocols are adhered without compromise.
The process can also be described as Examination – Evaluation – Education.
We look forward to any views pertaining to our Dia-Bese concept.

Sep 24, 2012

Fascinating discussion, and so glad to see Dr. George Lundberg weigh in here with his eminently common sense response.

As a dull-witted heart attack survivor, I’ve often thought that if only the E.R. doc who sent me home with a GERD misdiagnosis had just Googled my textbook MI symptoms that morning (chest pain, nausea, sweating and pain radiating down my left arm), I’m pretty sure that he and Dr. Google would have come upon the correct diagnosis – instead of pronouncing confidently: “You’re in the right demographic for acid reflux!”

Anybody who has read Dr. Jerome Groopman’s excellent book ‘How Doctors Think’ has experienced a crash course in the many personal decision-making biases that come into play when physicians try to navigate the medical mysteries of health care. Over-diagnosis, under-diagnosis, unnecessary procedures, marketing-based medicine, Big Pharma conflicts of interests – let’s face it, Khosla’a controversial stick-poking is a reality that’s been a long time coming.

Sep 7, 2012

The key in my eyes is not eliminating docs, but eliminating insurance co’s and the govt. Here in the US the trillions wasted every year to keep DC and the insurance co’s in the money is kiiling healthcare and our pocketbooks.

As for the future of tech in healthcare… only a fool would totally dismiss it. As for projections… nobody knows for sure.

Sep 6, 2012

Vinod Khosla is ahead of his time. We have succeeded on applying algorithm for Diabetes Complications Management.
I endorse this claim the concept that 80% of Doctor’s intervention can be replaced by technology, in addition to a well analysed data will give far better outcome that traditional approach.
Vinod’s point of view must be considered on case to case basis and not ridiculed just because some Doctor’s are trying to defend their narrow views and feel threatened or offended.


Seriously Nitin, it sounds like you think doctors are against effective healthcare transformation. This is not a battle between doctors and technologists, it is an opportunity for effective collaboration. However it has to be based on facts, not just hope. Medicine is a complex science, and there is great danger in oversimplifying it. Technology is also very complex, and I admit medicine has much to learn about it and from it. Some technologists, like some doctors as you have so rightly stated have narrow views and are easily offended. They are not the ones that will help transform healthcare.

Sep 7, 2012

I am so glad that there are Doctors who are open minded and willing to get some of their burden. Example could be of Symptoms Analysis and providing non-pharmacology therapies which enlightens the treating Physicians and the Patients ends up getting holistic support from a single source.
Over a period of time the gap will narrow down and technology (Devices + algorithm program) supported information will become a powerful tool. Busy Doctors who are supposed to go through numerous issues and time being their constraints, occasionally tends to compromises the optimum care their Patients deserved.


Vinod Khosla has triggered a very interesting discussion. Unfortunately he has taken some very large leaps of faith about how healthcare is delivered. First tests alone do not diagnose disease, they are a tool used by clinicians to support their clinical acumen. You need to first suspect a disease and order the relevant tests. As a general rule we don’t just do a random set of tests hoping to find an answer. So Khosla’s assumption that a machine can diagnose better than his so called “witch doctors” is both unproven and essentially wrong.
But it gets worse. Comments that doctors act on tradition and not science may have been correct in the middle ages, but all of my peers are very well versed in science. We call it evidence because it is science based on facts, not just assumptions based on watching too many Star Trek episodes.
Then there is the oft quoted statement that we need to focus on wellness not sickness. Unfortunately even healthy people get sick, whether it is as a result of a traumatic accident, or diseases like cancer, or auto immune disease, or infection. Sure we can use preventative strategies to keep people healthier, but ultimately they succumb to the vagaries of sickness. At that stage you really want a doctor trained in sickness to help you get better?
There is no doubt that technology will assist in making healthcare more efficient and effective. When it helps doctors are eager adopters. However slandering an entire profession you have no understanding of is not the best way of promoting it.

Sep 5, 2012

The debate should not be so focused on whether a machine will replace a doctor, or should replace a doctor, or could replace a doctor… but on whether a machine can make a doctor better. Take an IBM Watson-like supercomputer loaded with the complex algorithms and health information necessary for most diseases and conditions, pair it with a doctor who knows how to spot and interpret the subtle facial expressions and body language that only a human practitioner with years of experience is able to do, and you will much more quickly get to the right diagnosis and treatment plan than either entity could do on its own. The technology is already here and would significantly lessen healthcare cost and human suffering.

Sep 5, 2012

Most healthcare is selfcare. And with the internet and all the great medical info that is freely available there (as well as the junk), technology has already replaced a huge amount of the need to see a physician for most things. Besides, most human medical ailments are self-limiting if some physician does not get in the way and make them worse.

southern doc
Sep 7, 2012

“most human medical ailments are self-limiting if some physician does not get in the way and make them worse.”

But most ailments that a physician treats are not self-limited.

Walk into any oncology office and tell everyone in the waiting room that their ailments are “self-limited.”

You really don’t know what you’re talking about.

Johnathan Blaze
Sep 5, 2012

Knowing the way healthcare is now, chances are the industry will be able to suppress technologies that threaten to make them obsolete, even at the expense of patient health and advancement

Remember, medicine in this country is about one thing and one thing only — MONEY. Doctors have a huge entitlement complex and think they deserve gobs of money for short, routine visits & procedures. They also have strong lobbying groups in Washington to keep their salaries artificially inflated.

The medical industrial complex, composed primarily of doctors, hospitals, and pharm companies, is a corrupt, immortal behemoth that financially rapes the people, and the government.

Aug 18, 2014

I just had to comment on this. I am not a member of the medical community, but I take exception to what has been said about money being the only consideration in healthcare.

Have you read a statement from Medicare, Medicaid, Medi-Cal, or insurance company lately? The actual amount paid to a healthcare provider (doctor, hospital, etc.) is so small that I wonder how they exist at all. It shows me that Government and insurance companies have total control over what is paid and it isn’t what I thought it would be!

Sep 4, 2012

‘Invest in wellness, not sickness’

Yeah too bad there isn’t nearly as much money in that and the public health workforce which often returns by far the most impressive societal perspective ROI in terms of outcomes has really been hit hard since ’09:

“The recession has battered public health; across the country, local and state health departments have shed 52,200 jobs since 2009 . . . . . Forty-one percent of local health departments expect to make even more cuts this year, according to the National Association of County and City Health Officials.”

As for the insistence that Big Data with increased measurement and analysis will be the Holy Grail to solving healthcare inflation cost curve you only have to look at how that has played out in education since the passage of No Child Left Behind:

“Overall, the last 10 years have revealed that while Big Data can make our questions more sophisticated, it doesn’t necessarily lead to Big Answers. The push to improve scores has left behind traditional assessments that, research indicates, work better to gauge performance: classroom work and homework, teachers’ grades and quizzes, the opinions of students and parents about a school. In his recent book The Social Animal, conservative columnist and veteran education commentator David Brooks identifies this bias—to emphasize and reward what we can measure, and ignore the rest—as a key reason why technocratic promises in social policy have largely failed to materialize. Research, Brooks notes, shows that the key to success is more often found in realms that resist quantification—relationships, emotions, and social norms.

Even the godfather of standardized testing, the cognitive psychologist Robert Glaser, warned in 1987 about the dangers of placing too much emphasis on test scores. He called them “fallible and partial indicators of academic achievement” and warned that standardized tests would find it “extremely difficult to assess” the key skills people should gain from a good education: “resilience and courage in the face of stress, a sense of craft in our work, a commitment to justice and caring in our social relationships, a dedication to advancing the public good.”

What is occurring in educational policy is eerily similar to healthcare policy in the U.S. the past 10-15 years.

Sep 6, 2012

I think there are opportunities to use Big Data in ways yet not imagined. Right now we don’t have the data aggregated in a way that Big Data approaches will work – claims data is not accurate enough or even the right data, and clinical and related data is locked up in multiple siloes for each person. With genomic information available on a patient level and personalized medicine on the horizon, Big Data methodologies will be necessary to fully realize the potential, I believe.

Sep 6, 2012

Possibly but genomic information and personalized medicine ‘have been on the horizon’ for 15 years now and they aren’t going to do anything to bend the cost curve.

Sep 7, 2012

You may be right, MG. There was a recent review article in Health Data Management written by Elizabeth Gardner entitled “Changing the DNA of Clinical Care”, however, that gave some food for thought about what will be available in the foreseeable future. Whether it has an impact on costs is obviously debatable – as it is with any new technological advance. The link to the article: