Software's Final Frontier: A Conversation With Charles Simonyi

An extended version of an interview from the September Atlantic

Charles Simonyi sets off on a space tourist trip in April 2007. (Shamil Zhumatov/Reuters)

JF: You are a longtime pioneer at Microsoft and elsewhere. What was the basic idea that made you think you needed to take a different path in software development?

CS: It came from looking at the gap between what the hardware made possible and what was achieved by the software. If you look at the growth in hardware, it has been astounding. Of course everyone knows about Moore's Law and the exponential increase in computing capacity. The increase in storage capacity joined that, and the bandwidth between the computers also followed that exponential curve. And that created some incredible improvements in our lives, but if you look at it just with cold eyes, the gap has been widening between the possibilities and the realities. You must remember, too, when 2001 came out--

JF: Yes, it would've been 1969 or '70.

CS: I was just looking at the implied use of computers and displays in that film, and it was literally quite right, especially in the sequences when they were going to the moon. The displays in the cockpit there were very much like the displays that you see in the cockpits today. Hollywood seems to be ahead of the curve of showing the capabilities.

Also, if you look at game improvement, starting with Pong and going to modern games like Halo, the detail and the excitement that you get is just tremendous. We have more memory than we can shake a stick at, which is wonderful. At home I'm moving all my DVDs to disc. And all my photos -- not only the present photos, but the past photos. And my CDs--everything is going on a disc. And the displays! Remember when Bill Gates was building his house?

JF: Yes.

CS: Of course there was this dream that he would be showing art on his walls on displays. We were kind of envious: "Oh my God. He is so rich that he can have a display just for art." And in fact, in his house, in many places, there was a double wall. You had a walkway behind the walls to accommodate the displays, which were four or five feet deep and required heating and cooling and everything else. Today, displays--I've just been to a Rolling Stones concert and they had like a 100-foot display. And go to Times Square and there are skyscrapers whose only purpose is to carry a display on their windows.

This all has to be activated by software. We have some terrific new applications, and in that I include search engines Google and Bing, and, of course, the Microsoft Office suite. The potential is much, much greater than that. And it really has to do with involving knowledge or encoding knowledge in a deeper way. That's basically the reason I thought that a different path was needed.

JF: If the potential of software had been realized the way other potentials have been realized, how would our lives be different now?

CS: I think that the basic answer is that nobody would be doing routine, repetitive things. Think of how many times you go to a new Web site, and after you've formed that intention and made the decision, the rest is routine and repetitive. Look at health care and the incredible number of repetitive and routine actions you have to take as a patient.

Amazon has been applying a sort of rationalization and organization to a wider area of retail--but the elimination of routine activities from your life is the main thing we have missed. I think what will happen is that the concept of what's routine--and therefore avoidable--will expand. For example, making appointments, maintaining your calendar. Better software will definitely get into those areas.

JF: One other backward-looking question: As you try to explain the differential rates of progress in hardware versus software, is it a failure of effort or are the problems of software intrinsically harder?

CS: It's the latter. I sometimes call software "distilled complexity." Hardware is actually quite a bit simpler. It's this fairly simple logical divide. The same thing is true for printers. For example, I often show a slide that shows the inside of a teletype from the 1960s--you know if you look at an old James Bond movie and you can see the Telex machine making the noise in the corner. And if you look at those Telex machines, how complicated they were, and you look at a modern inkjet printer and open it up and there are practically no moving parts in it, you know there's the head, and there's the band that pulls the head along, and that's about it.

The other example I use is the Merganthaler linotype machine.

JF: Yes, I actually set hot lead on the college newspaper on one of those things.

CS: These were wonderful, wonderful machines. But, you know, they were expensive. They were dirty.

JF: And because they were complex, they broke, too.

CS: They broke, and that added to the expense of having to fix them. So what we've done now is, in a way, we've swept the room. We've swept the complexity into this thing, which is the software, and then we could afford to super-optimize the hardware.

JF: It's all sort of standard performance indices as opposed to basic architecture.

CS: Right. Nobody cares anymore because that's not where the action is. Certainly, for example, when you talk about memory--talk about simplicity! What's simpler than memory? There's an address and there's a bit there, and it had better get the bit back. It's either a 1 or a 0, and if I put a 1 there, I want a 1 back, and if I put a 0 there, I want a 0 back. And then do this as many times as you can, as fast as you can. That's it. End of specs. And that's why chip manufacturing moved completely out of the U.S. because it's--

JF: --Sort of purely execution.

CS: Execution, yeah.

JF: As you think of the software of right now, of 2013, which part of the software that people interact with are you most impressed by, for its incorporation of intelligence? Which one most annoys you for its lack of intelligence? And 10 years from now, what will be the most dramatic change for people?

CS: I think that just in terms of usefulness and the incorporation of intelligence, the simple task of spell-check is very impressive, actually. The statistical translations are getting surprisingly useful for purposes that people use them for. When you go to Wikipedia and misspell a name, for example, you will get very, very good suggestions, and you get there very quickly even from limited context and hits. I think that's amazing, and I'm in awe of that.

In terms of annoyance, I will tell you that what annoys me the most is consumer electronics. Consumer electronics is the most pathetic piece of--

JF: Crap?

CS: Your word!--that is foisted on us. The worst part of consumer electronics are the remotes. I go to a place, be it someone's home or a hotel room, and I'm faced with four remotes, none of which are labeled, and each one has about 50 buttons on it. None of those buttons have anything to do with anything I would like the display to be used for. Anyway, enough of the rant.

JF: Yes, I think everyone would recognize that. So is that where you think we'll see the most embedded intelligence in five or 10 years or in some other area?

CS: I think that it will be completely rationalized. The interesting thing about that is that there really doesn't need to be that much knowledge turned into it. It's covered by my phrase of "repeated activity": You don't want routine, repeated activity. I think that's where the intelligence will come in. It will be in health care and, generally, the organization of your life.

In some sense, everybody has a secretary available in terms of word processing and in terms of e-mail. We don't send faxes. We don't ask somebody to fax things. We don't ask somebody to create or send a letter. We can do it ourselves conveniently. But, for example, in terms of organizing our lives, we don't have a secretary who knows everything and can make obvious judgments without asking us, follow up on things--all of those tasks that a good executive secretary would be doing. I think that will be available in five or 10 years. It will be an app.

JF: God willing. I know that part of your team with David Allen is working toward this goal, right?

CS: Absolutely. Well, let me say a qualified yes. The cooperation with David Allen has to do with, of course, getting things done, which is a subset of that. You of course know everything about that. But it's a self-generated portion of your life--you still make the decisions. The human need is to be happy and be served, but the theoretical, technical need is to encode things in an intentional way so that computers can act on them.

JF: In various fields of expertise, whether it's aviation or genetics research or professional sports or art, there are things insiders know are difficult that people on the outside world don't recognize as being difficult. What is the underappreciated hard part of the work you and your company are trying to do now?

CS: I want to say facing up to complexity, but we don't do that. We are trying to express complexity. We have to leave behind some of the established patterns, but we have to kind of find this narrow sweet spot where we leave behind some of the patterns that got in the way as things evolved. So it's going forward in an uncertain path with a careful mix of renewal and conservation. We also retain those patterns that not only served us well in the past but will serve us well in the future.