The Internet will save higher education, but it may kill your alma mater.

I hated being a public company CEO.

I studied to be an architect. And I find tremendous similarities between building a company and the design process. Businesses have to do their planning on the fly in a fashion similar to an architect sketching.

Americans have learned to trust free markets. Republican or Democrat, we believe the unimpeded exchange of goods and services will yield better solutions than five-year plans set by even the most well-meaning public servants.

You still remember your SAT scores. And everybody else does too. Everybody's forgotten everything about themselves, everything else about high school. They remember their SAT scores.

Traditionally, universities have seen size as potentially dilutive to quality. If you doubled the size of campus and faculty, most would argue that you would make it a less compelling school. However, online schools will be as good as their classroom peers only if they are large enough to afford a substantial and ongoing investment.

Exciting discoveries in neuroscience are allowing us to fit educational methods to new understandings of how the brain develops.

I am always making sketches of how information should look or mapping out a marketing campaign. When I present my notes, people start responding to them. Desktop publishing makes everything look slick. When you present sketches, it helps start the dialogue and collaboration.

I played street hockey in Riverside Park when I was a kid. I played goalie. I didn't make the hockey team in college, so I played lacrosse instead. I didn't play hockey again for 20 to 25 years, and then my son became interested in the game. I decided to pick it up again. A friend let me play backup on his team.

We should get rid of the SAT as fast as we can. Look, there are bigger problems in society. This is not the biggest problem we have. But it's so easy to get rid of it. Right? Just pull the plug.

I'd never taken an SAT prep course. They were for weaker students.

The thing about startups is you can make it, and if it's wrong you can remake it, and you can build a team that you want to have, a product that you want to have. You're utterly focused on your users or your customers and their needs, and trying to figure out how to meet those needs.

The question was, in a sense, at Princeton Review, how much value was I adding as a public company CEO. I was adding less than other people might've... I think you want to move on when you've given your best work and then feel that you're not going to add as much value moving forward.

There shouldn't be one Sal Khan - there should be 100,000 of them.

After graduation, I took a job with Manufacturers Hanover Trust in software development. I don't think I was there more than a month.