In his first autobiography, Source Code: My Beginnings, Bill Gates recounts, at 13 years old, signing up for a 50-mile hike through the snowy Olympic mountains. Around the same time, he’d been introduced, through school, to his first computer, the PDP-8, with just 4 kilobytes of RAM. Despite the misery of the cruel December elements, he found the long spans of time coupled with the rhythm of his steps helped him parse, refine, and tighten lines and lines of computer code in his mind. That particular code would eventually be at the heart of BASIC, the software that put Microsoft on the map and helped birth the personal computing revolution.
The book details Gates’ family life and his upbringing as a smart but complicated kid fascinated by puzzles and programming. One who was fortunate enough to be coming of age at the same time engineers were discovering the first semiconductor chips and beginning to fulfill the promise of Moore’s Law. Gates spent much of his teenage years trying to find ways to get on computers—which weren’t easy to find back then—and write programs for them. Of course, all of this leads to the founding of the largest and most influential software company on the planet.
Source Code is out now and is the first of three autobiographies Gates plans to publish. The second book will focus on his years of creating and running Microsoft, while the third will tackle his philanthropic efforts and building the Gates Foundation. I found the first memoir to be an easy and enjoyable read: not only an interesting peek into the complicated mind of Gates, a brilliant tech pioneer, but also a fun trip back to the beginning for PC enthusiasts.
I sat down with Gates to talk about the book and catch up on some of the themes we discussed for PCMag’s 40th Anniversary in 2022. The following transcript of our conversation has been edited for length and clarity.
Wendy Sheehan Donnell: Source Code gives us a nice peek into your childhood, with fun, purposeful hobbies like hiking and card playing with your grandmother, a supportive family, as well as tight, like-minded friends, including Microsoft co-founder Paul Allen. You also describe yourself as a kid with excess energy who was “hyperkinetic, brainy, contrarian, and tempestuous” and that being something your parents struggled with at times. What about your upbringing and your personality makes you a success?
Bill Gates: I was extremely lucky to be born when I was. Because I’m a young person and getting this exposure to computers, which given that I liked puzzles where you could see whether you’re right or wrong, the idea that you write a program and it either works or it doesn’t work just fascinated me.
With my parents, I had this incredibly benign existence. I was allowed to go on hikes, I was allowed to be a page in Congress. And even my sneaking out at night—my parents weren’t overly concerned about that.
I was male, so my parents were just a little more open-minded about letting me do things. They were frustrated by the fact that teachers would sometimes say I was brilliant and sometimes say I was a real problem. There were times when I was very argumentative and sarcastic with them, which is why they sent me to a therapist to try and understand why I was fighting with them. Somehow, I saw that as a valuable way to put my IQ and energy in that direction. It took a year or a little more, but the therapist did manage to make me realize it was a Pyrrhic battle that wouldn’t help me in the long run, and fortunately got me thinking differently.
Bill Gates and Paul Allen at a Teletype Machine, Circa 1970 (Credit: Lakeside School)
Also, there’s my friend Paul Allen, who talks to me about chips. He read Electronics where, in an article deep in the magazine, Moore’s Law was first talked about. Paul said to me, ‘Hey, if this really does double every year or two, what does that mean?’ I said, ‘It means computing is free, and that is crazy stuff!’ That’s the point at which we realized we get to be in on this and maybe help it happen and lead parts of it.
Then my computer exposure starts writing BASIC programs. C-Cubed enlists us to try and crash their computer so they don’t have to pay rent, and that’s where we pull operating system listings out of the garbage can and learn machine language. Then we created the scheduling program for our high school, which was very complex. The best was when I took most of my high school senior year and worked on the digitization of the Bonneville Power Grid for defense contractor TRW. They had incredible people there who would review my code, tell me what was good, and tell me what wasn’t good. By the time I head off to Harvard at 17, I’ve had something like 4,000 hours of programming.
The big argument with Paul was him saying, ‘Okay, let’s make a personal computer.’ I’m like, ‘No, most of those things we’re not unique at, and IBM and Texas Instruments, they’ll do that better than we will. Let’s just do the software.’ Since that’s when computing is free, that’s going to be the limiting thing. Then, of course, the kit computer shows up, and we finally have to get going, or we feel like we’ll be left behind.
WSD: Right. The $400 Altair 8800 kit computer was made by MITS, where Dave Bunnell, the founder of PC Magazine, worked.
BG: Exactly, and I encouraged Dave in that, because Dave was super supportive. It was great when he started PC Magazine.
Small World: The Altair 8800 on the cover of Popular Electronics in Jan. 1975. Back then, the magazine was published by Ziff Davis, who went on to found PC Magazine in 1982. (Credit: PC Magazine)
WSD: The book details hours and hours of coding. Do you ever code anymore? Or is that something you dropped when you started running the company? Do you still dabble at all?
BG: The last Microsoft product that had a substantial amount of my code in it, sadly, is in 1985, when we made the M100, which is just a little 8-line LCD, a very portable thing. In fact, some reporters used it for a while because you could type in text, and it was just this nice LCD, battery powered device.
Thereafter, I do a lot of architecture, like in Excel or SQL, and I’m the one reviewing the code and saying, ‘this is too slow, or this is too big.’ But I didn’t write thousands of lines of code after 1985.
I still write some things today. Mathematica lets you do some symbolic things and look at statistics in a way that I’ll often play around with. When new languages come along, like Python or C#, I’ll write some code just to make sure I understand all those things.
But no, I’m embarrassed, I’ve gone from being an individual contributor to a muckety-muck manager of managers of managers.
WSD: That’s exactly what happens. I’m so excited to be interviewing you right now, because most of what I do these days is sit in meetings and strategize. I completely hear you on that.
Related question: You were a programmer, not a slick salesman, like, say, a Steve Jobs. In Microsoft’s early days how did you balance technical innovation and business strategy and actually selling the software? Because in the book, it seems like it was both you and Paul who are out there, and that seems like a hard skill when you’re kind of a nerd, right?
BG: Paul didn’t like selling, and he wasn’t involved much in that side of it. I had to hire those people and get out and do that.
The nice thing about software is, if people pay you, then you can write more software. It was kind of natural that if we got ahead, if our BASIC was the best one, that’s our first product. So I got Radio Shack, Apple, and Commodore to buy it: Those are the first three machines that are not kits. The Altair was for a pretty narrow group of hobbyists, although some people eventually got floppy disks and did serious things. These are the first computers, what we now call 8-bit machines, that you could really tell people, ‘okay, you can mess around and it’ll be fun.’
Then it’s not until 1981 that somebody makes a 16-bit machine, because the Intel processors had moved on, and that opened up the memory size. We were basically limited to 64k on those early machines, except for some hacks, the Intel 8088 20-bit address space, and so we’re 16 times better off. And then with the 386, wow, we’re really cooking, because then we have a 32-bit linear address space. We’re like a VAX computer, a really big, powerful machine.
Actually, the 68000 did the nice linear address space even before Intel did. The 286 was a weird detour that we told Intel was a mistake, but anyway, they did it, and it was important for a while.
WSD: You mention all of those machines, but what’s the first computer you purchased for yourself? I think the book probably cuts off before you could really buy a personal computer, but what was the first PC you owned?
I had a number of cool PCs I got to play around on: Incredible 8-bit machines from NEC, Hitachi, OKI, Matsushita, and others…I was able to extend BASIC in some pretty novel ways.
BG: Well, the book ends when I move to Seattle in 1979 and Microsoft has, I think, 18 employees at that time. I’ve already met Jobs and Wozniak. I’d already done AppleSoft BASIC, which was our 6502 BASIC, of course. I’ve already worked with Chuck Peddle on the Commodore PET. I’d been down and met John Roach, Steve Leininger, and Don French, who did the TRS-80. But I haven’t hired Steve Ballmer yet, and IBM has not made the PC yet.
I had a number of cool PCs I got to play around on with BASIC: incredible 8-bit machines from NEC, Hitachi, OKI, Matsushita, and other companies that people have never heard of. And it was those Japanese machines where I did things like build the Player, which allowed you to play music. Also the Grapher. I was able to extend BASIC in some pretty novel ways.
I put a lot of that into the software we did for those first 16-bit machines. Chuck Peddle leaves Commodore and makes the Victor. That’s actually the first 16-bit machine, and Digital Equipment made one, and Radio Shack made what was called the Color Computer. Weirdly, the IBM is what everyone thinks of, even though that wasn’t the first 16-bit machine, but everybody eventually adheres to the even BIOS specifics to have binary compatibility. Compaq shows the way on that for everyone.
I had a bunch of those 8-bit machines. I mostly worked day and night, so it wasn’t like I goofed around.
WSD: That leads into my next question, inspired by my 11-year-old: He wants to know your Xbox gamer score. For somebody who’s always been really into computers, you don’t talk about gaming much. I guess probably because you didn’t have much time to game. I’m also wondering if your parents would have allowed video games.
BG: I hope my mom would have limited my behavior because she was very on top of how I was spending my time. She was even a bit surprised how much I would disappear and go read. The early Atari games, like Breakout and Pac-Man, I was very good at. Compared to some of the sophisticated games we have today, they’re kind of a joke, although some people love the retro thing. I even taught my son to play Breakout and Pac-Man and some of those original, great things.
On the PDP-1, some people had made a game called Star Wars, which is this original thing where you had a trackball, and you could fire at little things. And so, the PDP-1 with this scope on it, was my first exposure to computer games. And in my dorm, Currier House at Harvard, that I revisited a couple of weeks ago, we had a Breakout machine in the basement that we wasted a lot of time on.
WSD: I read that you do play Wordle and the The New York Times Spelling Bee. Me, too. To me, that’s gaming, but again, my 11-year-old would say, ‘Not really gaming, mom!’
BG: No, it’s a funny type of game. I play Spelling Bee, Wordle, Quordle, Octordle, Connections. Those I do every day. It’s fun because I have friends that I’m competing with.
WSD: Me, too. I have a text thread with some of my family where everybody’s sharing their scores. I started playing Threads recently, which is also fun. I do the trifecta each morning before I get out of bed.
Back to business: What do you think was the pivotal moment when you knew Microsoft was going to be a major success? Did that happen in this book, or does it not come until later?
BG: I’m always a bit running scared. I had a lot of companies that went bankrupt and didn’t pay me. In the book, I had the issue where MITS is not giving us our royalties because of a contract dispute. And so, both Steve and I were always very conservative about things: Let’s not hire too many people because we really do want to be able to pay the salaries. I always said we needed to have enough money in the bank to pay everybody for a year, even if nobody paid us at all for the software.
I had seen Digital Equipment and Wang get decimated, even though I thought of them as amazing companies. Almost all those mini computer companies, Data General for one, there were a bunch of them. They’re gone. I wanted to avoid that. It’s not until the late 90s when Windows 95 and Office 95 are the highest market share in spreadsheet, word processing, database presentations, and we redefine everything as graphical interface and the integrated office concept. That’s when a lot of people hadn’t realized that Microsoft was so much of a software factory. We weren’t a single-product company.
Gates in the Charter Issue of PC Magazine in 1982 (PCMag)
WordPerfect, a great company, but they’re a single-product company. Novell, pretty much a single-product company, Lotus, although at the end they try and respond to Office by buying Ami Pro, they’re a single-product company. We were never a single-product company, and we were always global. We got a lot of revenue out of Japan even in the early years. In fact, in the 80s, there were a couple of years we got more revenue from Japan than we did from the United States, partly because of this amazing relationship I had with Kazuhiko Nishi who ran ASCII magazine, the equivalent of PC Magazine in Japan. He comes and finds me and tells me, ‘Oh, you’ve got to come to Japan,’ and he’s very helpful to our success.
But no, I was running scared, even though we were moving faster out in front and had a broader concept of what we were doing than any of our competitors. It’s not really until Google comes along with this idea of hiring massive amounts of smart people and doing all kinds of software products. We didn’t really have a full horizontal competitor until Google, and that’s more than a decade later.
Recommended by Our Editors
WSD: Speaking of, Microsoft turns 50 in April. That’s exciting. At this natural reflection point, what do you think the next half-century looks like for the company? I just came back from CES where Copilot is everywhere in everything. The AI PC is here.
BG: When I left Microsoft in 2008, I stepped down as CEO in 2000, I leave full time in 2008, I’m very apprehensive that things are going to get messed up, I won’t know what to do, and I’ll just feel bad about it. I feel very lucky that first Steve [Ballmer], and now Satya [Nadella] kept the company very much at the forefront. Not without some ups and downs, but it’s still really incredible.
I’m very lucky that Satya engages me in doing product reviews. I wrote memos about agents and higher-level semantics forever, and people were like, ‘why is he talking about that?’ Now that we have these AI reading and writing capabilities, they go back and look at those and go, ‘okay, yeah, we really do need to help people plan a trip as a broad thing. We really do need to organize by activity, not by the date that an email came in.’ The low-level semantics of the software, finally, everybody agrees can do something dramatic.
AI is the biggest change agent in the world. Even though it’s kind of a continuation of the whole digital revolution, and of course, completely built on it, it brings new challenges and new product categories. I’m thrilled that the OpenAI guys who I also get to work with are doing amazing things, and Satya has figured out a great alliance with them.
How do you put AI into Office? I’m pushing. They’re never moving as fast as I want, so I’m pushing hard about how Office can do even more with AI, and they have a very aggressive roadmap for that.
WSD: I know that you are very optimistic about AI helping to solve many of the world’s problems, but you’re also focused on combating climate change. How do you reconcile AI’s impact on the environment?
BG: Well, my theory of climate change is to innovate so that the clean, green approaches cost either the same or less. Honestly, this demand for electricity, I hope that it takes us to the next level in terms of nuclear fission and fusion and doing the grid in a better way. Also figuring out energy storage so that renewables can be a big part of the mix. It’s good that the big tech companies are still willing to get involved in early-stage things like nuclear and help bootstrap them because they care about climate and they’re willing to pay a bit of a green premium for a period of time so that they’re part of the solution.
The actual demand from data centers, although it’s urgent, is less electricity than if we electrify every car. It’s less than if we change all-natural gas heating in a household to electric pump heating. And so, we were already facing, before this data center realization, the need to double the electricity output in order to accommodate electric cars, electric heating, and electric industrial processes like cement and steel. This is after 30 years of flat electricity demand. Eventually we can do things like solar power and actually get the cost to be quite economic. Although the first few plants will need the customers and the US government to help out. Because if you take nuclear, you have to build about 20 plants before you can compete against natural gas.
WSD: When I interviewed you for PCMag’s 40th anniversary in 2022, you said, ‘With AI, there’s huge potential to do good in the world, but it’s also important to be cautious and make sure it’s being developed in the right way.’ Three years later, how do you think that’s going?
BG: Well, I don’t think our understanding of how we avoid turmoil because AI is so powerful, we certainly don’t have the answers to that. I think the discussion, even at the level of concern, is good and there’s a broader audience. I was pretty stunned in the 2024 election by how little AI came up. I think in the 2028 election, the biggest issue will be AI policies, where each party will have different views on how we take advantage of the good things and avoid the problems we have there. I do think that books like Mustafa Suleyman’s The Coming Wave or Harari’s Nexus should be widely read.
People could say this is not the first time that when you have a new technology, it might be used to ill as much or more as it’s used to gain. It can be used in weapons. I love where in his book Harari explains that when the printing press comes along, there are more books written about how you find and kill witches than there are about the scientific revolution. And so we map our human strengths and weaknesses onto any new tools we get, and AI is the scariest one of all.
We map our human strengths and weaknesses onto any new tools we get, and AI is the scariest one of all.
WSD: Exactly. First, it was computers, then the internet, and AI is the next revolution, right?
BG: Yeah. I mean, computers and the internet felt very benign and mostly empowering. In fact, when people asked what problems do you see, the main thing we would talk about is the digital divide. This stuff is so good, we gotta get to get it to the inner city and the countries in Africa. Not, ‘oh my God, it might make crazy people no longer believe in facts and they could find each other and reinforce their non-factual beliefs.’ I think I was naive about it: Eventually with things like social networks and AI, we’d have to shape the use of the technologies to have the good outweigh the bad.
WSD: 100 percent. I guess my last question is, you’ll be 70 this year, you’re running a foundation, you’re still very involved in Microsoft, you’ve got two more books in the works. Do you ever want to retire?
BG: It is strange how much I enjoy my work. Obviously, I’m not working for financial reasons. I’m still traveling a great deal, but it’s because my work is fun and fulfilling. In the Gates Foundation, we’re trying to end polio. We’re trying to cure HIV. I get to work with brilliant people. The Breakthrough Energy stuff, the number of new ideas that we’ve been able to back there has been way greater than I expected, and we have a fantastic team.
And all the time I get with Microsoft, not only am I able to help a little bit, but it informs me how quickly things are moving. The Gates Foundation is figuring out, ‘okay, how do we help education in the inner city? How do we help the doctor shortage in Africa? How do we help farmers?’ We’re trying to give African farmers better weather forecasts than rich farmers get, better farming advice than rich farmers get, and that’s using AI. The time with Microsoft makes me well equipped to lead the way to using AI for the values and the goals of the foundation, which are very much about helping the poorest countries.
Get Our Best Stories!
This newsletter may contain advertising, deals, or affiliate links.
By clicking the button, you confirm you are 16+ and agree to our
Terms of Use and
Privacy Policy.
You may unsubscribe from the newsletters at any time.
About Wendy Sheehan Donnell
Editor-in-Chief, PCMag / VP of Content, Ziff Davis
