The cost of adding a feature isn't just the time it takes to code it. The cost also includes the addition of an obstacle to future expansion. The trick is to pick the features that don't fight each other.

The core of what I do is solve problems, whether that's in graphic engine flow or rockets. I like working on things that are going to have an impact one way or the other.

One of the reasons microcomputers progressed so fast is people are willing to accept crashes. It's faster to build something and try it, even if it means you'll have to rebuild later... If you spent too much time building and massaging one vehicle, you don't learn anything.

The great games are the space sims and driving sims and these experiences where you're basically sitting at a table with nothing happening in front of you. A lot of interesting things are evolving there. There are great games that can be made.

The truth is, you cannot run a political campaign like a tech startup. Technology is a field that fetishizes disruption. The old ways are suspect, and we place an almost irrational trust on new tools. That's fine for developing games, but it was a failing playbook for politics.

In 1962, we created the Filmmakers' Co-Op because nobody wanted to distribute our films. If we had the Internet in those days, we wouldn't have needed the Co-Op.

The only technology that can 'see' beneath the ground is radar imagery. But satellite imagery also allows scientists to map short- and long-term changes to the Earth's surface. Buried archaeological remains affect the overlying vegetation, soils and even water in different ways, depending on the landscapes you're examining.

I'm amazed at the amount of time people spend on the Internet.