Sun Microsystems co-founder Vinod Khosla explains entrepreneurship and motivation

Sharing wisdom.
Sharing wisdom.
Image: AP Photo/John Raoux
We may earn a commission from links on this page.

Long before the current wave of entrepreneurs and startups hit India, Indian Institute of Technology-Delhi graduate Vinod Khosla made it big in Silicon Valley by co-founding computer software and hardware firm Sun Microsystems in 1982. The company was acquired by Oracle Corp in 2010 for $7.4 billion.

Khosla now runs a venture capital fund—Khosla Ventures—which invests in companies involved with computing, mobile, silicon technology and clean technology sectors. In 2014, Khosla was included in Forbes’ “Midas List,” which claims to name the world’s smartest technology investors.

On Dec. 2, Khosla took to question-and-answer website Quora to shed light on technology and entrepreneurship. Here are edited excerpts from his insights:

After amassing so much wealth and fortune, what keeps you motivated?

Wealth is only important because it gives you the freedom to do what you want to do. Wealth isn’t an end goal. Hence, you don’t need wealth to be happy. But it does enable you to do many things that you might be interested in.

I like to say you grow old when you retire, not that you retire when you grow old. The key is to be motivated and excited about what you do and I’m so excited about helping entrepreneurs with so many innovations the world needs and are fun to do that I have way more ideas than the time that I want to work on. That’s what keeps me motivated.

If you were to start a company today, what would it be? Why?

There’s [sic] a few large problems in the world that I’d love to solve. The biggest long-term problem may be climate change, but I can’t think of one company I could start that would solve the problem. So investing in lots of different approaches is right for it.

I think, education is a big problem area but that’s best addressed iteratively by a lot of people trying lots of different things.

There are other problems but the one that’s most amenable to technology solutions—and the one I would probably address—is completely automating the practice of medicine, so everybody everywhere has the highest quality care through their cellphones.

This is a machine learning problem for which most of the data already exists.

Where will the next ”Silicon Valley” be?

To me, Silicon Valley is not a place, it’s more a state of mind and a culture of how things are done. It’s a culture in which people are willing to bypass convention in any area, not be overly biased about why things can’t be done, but rather take the approach of how one might take a shot at it?

It is one where people are not afraid of failure but just look at the consequences of success—not the probability of failure. This culture can be replicated in any part of the world, but is discouraged in most places, partly because failure is discouraged and partly because planning predictability and success along predefined pathways is the only way things get funded instead of saying something is just worth attempting.

Having said that, there are ecosystems around places like Bengaluru and Israel that are starting this transformation of culture. More established places like Europe see less of this, unfortunately, even though they have a lot of talent to make this approach possible.

CEOs with a business background usually fail in leading tech companies, as opposed to engineers. Do you agree with this?

Reasonable people do reasonable things and only unreasonable people can do unreasonable things. Most practical business people are reasonable. Building a startup is so hard with so many obstacles and ups and downs in the way that practical people become too pragmatic and turn big visions into decent ideas instead of sticking with their original vision.

It takes more than most people with normal business backgrounds—it takes people with unreasonable visions to build unreasonably successful tech companies. There are exceptions, but between the two kinds of CEOs I’d rather have somebody who is naive but with a great vision and understanding of what technical innovations could enable and can look at the problem by imagining the possible instead of putting together a series of pragmatic steps—which business people do.

Why haven’t medicine and computer science made more progress to automate clinical diagnosis?

Disruption and revolution are fun words in the technology world but the receivers of disruption and revolution usually get hurt. In any large disruption, the people who get disrupted always resist the disruption. Taxi drivers will oppose Uber even if it’s a substantially better system. The music industry will resist a democratisation of music and healthcare systems—especially in the US—will resist disruption of that system.

Beyond just the economics of disruption and disrupted there is generally a lethargy in most systems in most areas to keep doing things the way they have been done because the familiar is what humans prefer. This is why many revolutions that should happen don’t happen and many take much longer.

Is Elon Musk right to say that artificial intelligence is our biggest existential threat ?

Many technologies have both good and bad aspects that they can be used for. Nuclear is the most visible example of this with nuclear bombs and nuclear power.

Machine Learning is similar, but probably more manageable. So I think Elon’s concern is valid and one we should pay attention to it. But it is probably more manageable than he might intimate. I can think of more scary technologies that we’re using today—DNA editing tool CRISPR being one in biology—which has equally scary potential. So while we don’t want to risk large catastrophic possibilities, I personally believe technical progress should not be stopped for fear of a downside outcome.

Having said that, we are years away from even remotely approaching the point where we would either be able to create that level of artificial intelligence or be subject to them going wild. Good debate to start, but a lot of work before we get to that point, and we don’t know if we could get there in 20, 30, or 50 years. But it’s unlikely to be in 5 years.

How will work and jobs change over the next 10-20 years?

The basic assumption is that all human judgement can be replaced. From picking weeds in a corn field to composing the best music, as surprising as it seems. That will then cause a real shift in what humans can compete at and we will get a dramatic increase in the number of things machines can replace humans at—the easiest ones being things like radiologists.

This also extends to low-scale jobs like making hamburgers, mid-scale jobs like driving and legal research, high-scale jobs such as medical decision making that doctors do, and extremely creativity jobs like composing music and creative writing.

If all these can be done by machines, what is there for humans to do? We’ll have great abundance and rapid GDP growth and rapid increases in unemployment or replacement of current jobs. Most people will only work on things they want to work on not because they need to work and that will be the future of work. Of course, we will have the problem of income inequality, which I cover in my piece.

These replies first appeared on Quora. The full version of the Q&A can be found here.