I’m going to post a lengthy excerpt here from the conversation I had with Bill Gates earlier this week, because, well, I have so much of it, and it kind of works as this wonderfully absorbing dramatic monologue about where Microsoft came from. This is him essentially telling the story of how he and Paul Allen figured out that writing software for personal computers was maybe a good business to be in. Which is pretty impressive since at the time, in the mid-1970’s, almost nobody had a personal computer.
He also talks about his math career at Harvard, his anti-social-ness, the origin of Steve Ballmer, his C+ in organic chemistry, why he was never a hardware guy, and other interesting digressions. (Note that even though he digresses a lot, it always makes sense — Gates always closes the parenthesis.) It’s amazing how excited about this stuff he still is; also interesting is how much credit he gives to Paul Allen. This is raw transcript, so Gates won’t always speak in complete sentences, but you get the idea. If you’re curious, the actual audio is here.
ME: So did you guys know, really know, that you had something special going on?
GATES: Sort of yes. It’s a very unusual thing, because I had…my obsession with the computers, and my relationship with Paul was at Lakeside, starting in eighth grade. And so we had done programs, we’d done the school scheduling, we’d sold a payroll program. We’d done a lot, enough that we took part of my high school time off doing a job for TRW. I helped Paul get a job at Honeywell, so he could be back there so we could brainstorm, because his deal was the microprocessor—we’d agreed the microprocessor was going to change the world, it was weird that people didn’t see that. He wanted to be back there to kind of convince me that, hey, we should do something about it. So anyway he was living just a few miles away, and that was my sophomore year that he came out and we were talking about those things.
It was interesting there. I thought of myself as a math person. And one of my classes freshman year was the one where everybody is the best math person they’ve ever met. And their personal positioning is, I am the best math person. We all had 800s on our SAT, 5s on our AP. And so we go into this thing. I’m having dinner, the three of us who came in best in that class, one is a lawyer in New York, super successful at Hughes Hubbard; one is a professor of mathematics and chaos theory at Cornell, Jim Sethna; and then I. So the three of us sort of survived out of the 80 people who had this personal positioning, we survived somewhat intact, and that actually made us very good friends.
But, anyway, I remember talking to one of these math guys, and he said, geez, this computer stuff, you know, you’re really good at that. I said, yes, but the courses here are so easy, and so I never signed up. And the only computer science course that I ever signed up for was the one that had the most prerequisites in the whole catalogue, and I signed up for the second half of the year, which is this thing on stochastic scheduling. So my freshman year, I show up and it’s all graduate students, and two days into the course I tell the professor, hey, you know, this thing is wrong, because I was a little — my social skills weren’t that great. So I told this guy, this queuing theory thing he did was wrong.
Anyway, so I was going to these super hard math classes, and staying away from computer science classes, partly because I’d done so much computer science, and I was doing computer science projects using the computer there. Paul and I did the microprocessor emulation on the big computer that was there, and that’s actually where we did the first version of Basic that he flew back with this paper tape to take to those guys [i.e., the guys who made the Altair, the MITS guys. But that comes later – ed.]
When I was in high school, we had bought a 8008 chip, which is the 1971 chip. But the 1974 chip, which is the 8080, which is the one where Paul hands it to me and says, look, this is the one — Bill, you’ve got to admit this one is good enough, because it was better than the PDP-8, which was a minicomputer that I had done a Basic interpreter for that we’d had on loan out at the high school. And Paul was right. The 8008 was way below the PDP-8, meaning the minicomputers there, and the 8080 was better than the PDP-8. So then the idea that you could do this incredible computer around it, he was certainly right. So we wrote that emulator.
But, anyway, the math guys were always saying, hey, you’re going to end up in computer science. And I said, yes, I think maybe so, but it’s not the courses — the courses weren’t that interesting. And the great thing was, if you were a major in what was called applied mathematics, if you wanted to get into any class, like an economics class that was overcrowded, or a history class, you could say you were applying mathematics to that subject, and so you got the privileges of that major, but they never kept track that you were pretending to be applying mathematics to economics one day, and then the next day applying it to history. So I had this wild card major that I could use, and it was key. That was how I met Steve, was I got into this graduate economics course that we didn’t have any of the prerequisites for, which was a fantastic microeconomics course taught by Mike Spence, who is — what is he now, head of Stanford Business School, or something like that?
ME: You know, the stories that people tell, and even some of the way you played it in the Harvard speech, is that, you know, you weren’t that serious of a student.
GATES: Yes and no. The whole thing where it was so many people, where you come from a small high school, and then there’s so many people, and they’re all so talented, it’s a little bit of a shock. I had a group that amongst ourselves we were quite social, but we were the anti-social group. And so we’d hang around and talk about how our courses didn’t really matter, and stuff like that. And I had this thing where after about three weeks in my freshman year, I would go to the courses that I wasn’t signed up for, and not go to the ones I was signed up for. Then during reading period, I would study. I was very hard core, I was at Hilles Library…first thing in the morning, and I stayed until the end of the night, and I studied like mad.
And so, I only got one bad grade the whole time I was there, which was in organic chemistry. I almost put it in the speech. I got so intimidated by the premeds, and so I’d go to watch the videotapes, and half the videotapes they had the video but not the audio, and the other half they had the audio but not the video, which in organic chemistry is terrible, because the guy will say, we take this chemical, and you hear the chalk going, and this chemical…so the audio is no help. I got a C+ in that course.
ME: Why were you even taking that?
GATES: Well, I love science, and my favorite course in high school had been in organic chemistry. You were supposed to take Chem10, which is inorganic, and then take organic. But I knew [inorganic], because I had an incredible chemistry teacher in high school. In fact, I was remembering, there is this book by Sacks called Uncle Titanium, which is about the fascination of some chemicals are shiny, and some aren’t, some precipitate, some don’t. [I only realized later that meant Oliver Sacks’s Uncle Tungsten. Yes, this proves it: I’m smarter than Bill Gates.] Anyway, that whole thing of the elements, and how they work. So I loved inorganic, but then I didn’t like organic. There are certain ironies to that, because now in my foundation work, I am spending all sorts of time reading medicine, and biology, and stuff like that, but that’s my C+ there. So I was a weird student.
This friend and I had this thing where when we would go down and sit at the table, we’d always discuss the courses we took the year before, and everybody else was like, why are you discussing those? We’re paying attention to these. And we’d like be going over the homework we didn’t turn in, or something like that.
The great thing about Harvard was, you could always bet it all on the final, virtually every course, and so I would just do this thing during reading period. And this friend of mine decided to do it the same way, actually Monte Davidoff, who helped us write the Basic, he’s actually the third guy who wrote a small part of the Basic. The first Basic is Paul and I, and then the floating point is this guy Monte. So Monte tried to imitate my thing where you don’t do much during the normal year, and then you study hard at the end. But he just freaked out. It really is hard on you. And so he got not very good grades. I decided I wasn’t going to tempt other people into this sort of perpetual procrastination, and actually when I got going at Microsoft, it took me a while to get out of that mode that the coolest thing was to wait until the last minute to get something done. That was my showoff kind of thing.
And so I was really a good student. I mean, like in that EC 2010 course that Steve and I took together, you know, I actually got the top grade on the final, at the end we had studied so much, so much, because it was — I still wanted to be a good student.
ME: Let me get back to that question I asked before. Everybody tells the story about that article in Popular Mechanics with the Altair on the cover. You’d grasped that this was an innovation of enormous importance. How much did you know about what was going to come?
GATES: Well, the key thing was, in 1971, in Electronics Magazine the 8008 comes out, and it’s like on page 94 of Electronics Magazine. So Paul Allen is reading Electronics Magazine. He was always much more of a hardware guy than I was. I could never — it was always these wire-wrapped boards you would do, and I hated that kind of stuff. But he showed that to me, and he said, look, with this doubling, with this doubling every year thing, this thing is going to be amazing. And I thought, God, he’s right. Our friends at Digital Equipment must be stunned, the microprocessor is going to change the rules.
And of course nothing happens in 1972, 1973, and then even Intel a little bit didn’t realize what they had. There were people coming to them like Bubcom, Datapoint and saying, hey, your chip can do more than just control elevators. Datapoint did some really cool stuff with it. Bubcom is a Japanese calculator company. And then when they come out with the 8080, even then they don’t know what it is.
So Paul had really gotten us talking about the idea of a machine that an individual can buy. There was a guy named Ted Nelson, who was writing some crazy stuff about computers and how neat and cool they were, and Bob Albrecht had done using minicomputers sort of a time-sharing thing called People’s Computer Company. So there were tiny hints — there were a couple of guys in France did a thing called MCM — of people trying to get computers down to play with. Now, our idea of playing with a computer was to sneak in at night, like at the University of Washington, and get the thing at the Department of Physics that nobody was using, and use it all night, because they were big and expensive, that was our only way. And our whole high school years, this thing of finding free computer time, which a university like UW, the number of computers up there that are sort of half-used is very large, the medical area, physics, and so those were our highlights of our high school years were sort of bumming computer time.
ME: Isn’t there some story where you scammed a little too much time and you got kicked off?
GATES: Yes, we did this thing where we proved you could steal the password file on this PDP-10, and then Paul and I were banned from using the computer for a year. Then six months into it I find out that Paul has found a computer up at electrical engineering, and has been using it without even telling me. He’d been using it the whole summer, and the whole three months. It was a great computer. So then I was like, Paul, why didn’t you tell me? And Paul actually was better than I was at finding things up on the campus. There was one we got into physics at night, the EE one was the ninth grade one. There was one down in the medical center.
So this idea that computers would be a neat thing to have, and that you could do some interesting things…literally, we’d written Monopoly game programs, and recipe card programs, all sorts of goofy stuff. So when we saw that magazine it was the thing that Paul and I had been talking about happening was happening, and it was happening possibly without us.
So in a way we were sad, it was a cold, Boston winter…it’s December that it’s there on that magazine stand and we buy it. And we’re sitting there going, oh no! It’s happening without us! So that’s when we call them, from the dorm, and say, hey, we can give you a Basic interpreter. We can license you a Basic interpreter to really make people able to use this thing. And they said, that sounds interesting. So then we went to work, and four weeks later we had the first version, which we used this brilliant simulation technique.
Then I called them up and said, hey, how do you get characters, if you connect a teletype up, what’s the I/O port, and the character-ready bit mask to get characters in and out of this thing? And the guy on the other end, his name was Bill Yates, he was the real designer of the thing, although Ed Roberts is the name everybody knows, said, that’s interesting. You must really have some software, because nobody has ever asked us before how you get characters in and out of the thing. Let me look it up and call you back.
So he calls me back and then two weeks later Paul flies out with the paper tape, which I stayed up all night checking to see that the book that describes this processor, that we read it right. If we just had one little thing about how you include instructions wrong he was going to fly out there, this paper tape would load up and nothing would happen. Nothing. We had to have everything right. So it was such a miracle.
They had never had a machine with 4k of memory before. So [Allen] writes the bootstrap loader on the plane coming out, loads it in, and the thing runs, and it runs fast, because our simulator was fairly inefficient. So Paul calls me up and says, damn, this thing works!