A few of you may have read my review of William Hertlings’ Avogadro Corp on Silicon Florist earlier this year, and if so, you’ll know I’m a fan of the book. I’ve also had the opportunity to meet William for coffee several times lately, and thought it would be fun to do a little interview with him, as he has a second book coming out, and you all should know about it too. Everyone, meet William. William, everyone.
Q: You’ve previously told me you were influenced by my talk Build Something, Build Anything. Can you tell me a little bit about that?
A: When I first heard you talk, it really resonated with me. There’s an example that Clay Shirky cites, which I think you mentioned, that at the time, Wikipedia took about 100 million hours of effort to create, which sounds like a lot of volunteer effort. And then he points out that Americans watch 100 million hours of TV commercials each weekend. Not even TV content itself, just the advertisements. We could be doing so much with our lives.
Q: What was the first something you started building?
A: In 2000 I worked on a photo sharing web application. I eventually quit working on the project, but it taught me some valuable lessons about the scope of what I could address by myself and what I needed to get help with.
But it was even more interesting in terms of what it was doing for my career. At the time I was on a team of six engineers at HP developing a J2EE web app. And the whole time, I was working on this project on the side. I’d figure out how to implement user accounts for my photoweb project, and then we’d be implementing user accounts at work. I’d figure out how to do database backed session management on my project, and then we’d be doing it at work. I felt like a rock star.
Building something doesn’t mean that everything you build will be successful by itself, but it teaches you skills you can use elsewhere.
Q: You had a similar experience with the Netflix Prize. Can you talk about that?
A: Gene Kim and I have been buddies since grad school, and we worked together at Tripwire back in 1997. I think it was in 2007 or so, he asked me if I would enter the Netflix Prize with him. That was the contest Netflix created in which they offered a one million dollar prize for anyone who could beat their recommendation algorithm by 10%.
For about six weeks I totally immersed myself in the Netflix Prize and learned a ton about recommendation engines and big data sets. After that we knew we weren’t quite good enough to win the competition.
But even though we didn’t think we could win the prize, we had learned enough to be able to do something useful with recommendation engines. So I just kept asking myself what would be the best project.
At HP, I’d been working on improving the customer support experience. One problem we had at the time was the difficulty for customers to find the right support document for technical issues. We knew that if we could do a better job of matching up customers to the documents they needed, more people would be able to solve their problems on their own, and be that much more satisfied.
So inspired by my work on the Netflix Prize, we built a recommendation engine around support content. We looked at the clickstreams of previous visitors, and determined what pages they viewed and what pages they were ultimately helped by. With this data, we could recommend, given a particular customer on a particular web page, the web page they were most likely to be helped by. These recommendations made customers 25% more likely to be able to get the help they needed on their own. It helped HP avoid expensive customer support calls, and helped the customer get their problem solved faster. Win for them, and win for HP.
Q: So that’s another case of a personal project inspiring something at work. But given that your first book, Avogadro Corp is all about recommendation engines, I’m guessing it was also inspired by the Netflix Prize work.
Yes, but like all of these things, indirectly.
After the Netflix Prize, I had an idea for another personal project: a wishlist integrated with Facebook. My idea was that you could combine people’s wishlists and profile data and the power of a recommendation engine to be able to make gift recommendations for someone who didn’t have a wishlist.
One lesson I had learned from my first solo web project was when to look for help. In this case, I asked my friend Nathan Rutman to work on it with me. We taught ourselves Ruby on Rails, and got coding. I want to give a shout out to the PDX Ruby Brigade, because they are always so welcoming and helpful to newbie Rails programmers.
But getting back to the book: Around this same time, I had read and was really inspired by Ray Kurzweil’s The Singularity Is Near. I drank the Kurzweil Kool-Aid and became a believer in Kurzweil’s predictions that strong AI was an inevitable outcome of exponential growth in computing power.
Then one day, sometime in the spring of 2009, Nathan and I were having a lunch meeting to discuss our wishlist project, and I was spouting off more Kurzweil Kool-aid. Nathan challenged me to come up with a logical, grounded explanation of how strong AI could come to be.
Based on our experience with recommendation algorithms, I came up with an idea on the spot. My example was around a bunch of Google engineers who decide to write a language optimization engine for Gmail that optimize the particular words chosen based on the intended recipient.
By the end of lunch I had filled a few napkins with notes and diagrams, convinced Nathan that strong AI was plausible, and convinced myself that the idea would make a great book.
Of course, Nathan and I had just embarked on this wishlist app, so I didn’t have a chance to write the book. But I’d think about it literally every single night while falling asleep. By the time we were done with the wishlist app in November, I’d spent most of a year thinking about the book, but written exactly zero words.
Then in December 2009, I took four weeks vacation from HP and wrote the entire first draft. On New Year’s Eve, my wife kept calling me: “Are you coming up? It’s almost midnight.” And I finished the manuscript a few minutes before midnight.
The wishlist app never really amounted to anything, but it provided the inspiration for what’s become a three book series and also led to my current job inside HP where I’m doing Rails development.
Q: You’ve told me that writing the books then inspired other business ideas. Can you tell us about that?
After ten months of trying to find a publisher for my first book, I decided to self-publish.
It was easy enough to create the e-book version – it was a few hours work to get a Word document uploaded to Kindle Direct Publishing on Amazon and formatted correctly.
But it took weeks of research and work on my part to design the interior layout of the print version. Everything from font family, font size, leading, margins and more are based on expert knowledge and are different for every book form factor. And it’s sad to say, but styles still don’t work correctly in word processors.
After about forty hours of work trying to format it myself, I eventually turned it over to a professional, Maureen Gately.
Based on the publicly available data I could find, about 750,000 books are self-published each year, but only 50,000 or so are made available in print editions. My guess is that the largest reason for that is because of the manual design work involved in creating those print editions. I think the right technology could eliminate all the pain of doing interior book design.
HP is expert in making stuff look good on the printed page, and we have a ton of technology to accomplish that. So this is a perfect opportunity for HP.
Q: I loved how Avogadro Corp is set in Portland, the image of a mammoth 10,000 employee tech company located in downtown, and the maps you shared showing Avogadro superimposed over what is now a trucking company. Why did you set it in Portland?
A: It started out set in Silicon Valley. But I didn’t know Silicon Valley well enough to write at the level of detail I wanted to write.
I know Portland. I’ve lived here since ‘95. Although I work from home often and from HP’s campus in Vancouver, my time working downtown at Tripwire in the nineties was very vivid. So I knew that Avogadro was going to have to fit somewhere in the heart of downtown. Then I started doing employee space calculations and realized I needed to find a pretty large spot. I fired up Google Maps and the obvious spot was the Conway Trucking site, which is about six blocks in Northwest.
For my second book, A.I. Apocalypse, I set parts of it in Portland, but I heavily used scenes from Brooklyn, where I grew up, and South Shore High School, where I went to school. There’s a few bits in Milford, Pennsylvania, where my dad and I built a house in the 1980s.
My third book is largely set in Tucson, where I went to grad school for three years.
Q: What can you tell us about the third book?
It’s called The Last Firewall, and it’s about a nineteen year old girl who is all that stands between an unstable AI and its quest for world domination. It’s kind of a cross between Buffy the Vampire Slayer and The Matrix.
It’s part of the same universe as the first two books, but it’s set another ten years out in the future. I try very hard to extrapolate out technology trends and try to make it as realistic as possible. So building on the first two books, there are intelligent artificial intelligences and robots, and they’re pretty commonplace. But what’s different is that by 2035, when the third book is set, computing technology has shrunk to the point where pretty much all people have neural implants so that they can be connected to the Internet 24/7.
Q: Any last words of advice for people who are setting out to build something?
A: It’s a given that you have to pick something you are passionate about. It’s tough to come home from a day job, and maybe also then take care of family and kids, and then still come up with the energy to do something else. So first and foremost, follow what you’re passionate about.
And then don’t worry about making money with your projects. I mean, money is always nice. But nine times out of ten, these things don’t turn out to make money. But they can lead to your next job, or the next project, or an entirely different field.
My day job is at HP, which is a big company and it’s almost the opposite of a startup environment. But in the last ten years, all of which was spent at HP, some of what I’ve done has been: web programming, customer support, project management, web based support systems, user experience design, data analysis, recommendation engines, social media strategy, social apps, more web programming, blogging, and writing three books. Luke Kanies, the CEO of Puppet Labs, jokes that every time we see each other I’m doing something totally different.
All of that has been really fun, a little bit profitable, and it’s given me tremendous flexibility when it comes to looking for jobs. None of that would be possible if I didn’t have a build something, build anything mentality.
Q: So, you mention that you’ve been doing side projects for around 10 years, and now you’ve really come to see how they can contribute to your full-time job, and other side projects. What was your ambition for your first side project when you started it, and how has that ambition changed? Was there a moment when it had the most dramatic shift?
A: With the Photoweb project, I had two main ambitions. One was to be able to earn enough to be able to work for myself, as opposed to a full-time job working for someone else. The second ambition was to influence the world in some beneficial way that was bigger than just myself.
Twelve years later, I think these are still my two main ambitions, but the change is that I’m tackling the money piece in a different way. Instead of looking for all or nothing projects that would have a big win (or none at all), I’m looking at projects that are going to generate some trickle of income. Hopefully it’s sizeable, but if not, that’s OK too. I think that shift in thinking was inspired largely by Tim Ferriss’s book The 4-Hour Workweek.
I’ve been blogging since 2002, and starting last year, I finally saw a steady income stream from my blog. Then as I’ve published my books, they’ve created another income stream, and that is increasing month by month.
I’m not going to be quitting my job yet, but I could envision a point in two or three years where it might be possible.
Q: You’ve mentioned that in many of your more recent projects, you’ve collaborated with other people more (especially the non-book projects). Have those people found the same value in side projects that you have? Meaning, do they share your same enthusiasm for them, even if they don’t turn out to be monetarily beneficial?
A: When I worked with Gene Kim on the Netflix Prize, I think the enjoyment and benefit was largely that we were working together. We could have been out at a bar drinking, and had a great time, but we could also code side-by-side, and have a great time. Creation is rewarding in and of itself, and working with a friend is rewarding in and of itself. I found the same thing working with my friend Nathan Rutman on the wishlist project. In that respect, I do think they shared the same enthusiasm.
That being said, I think you need to have a large number of projects going before you start to get synergistic effects where one project ties into another project. If you’re narrowly locked into a day job and you’ve allowed the scope of that job to be defined by someone else, then you might not be able to use what you’ve learned elsewhere. I think having a positive feedback effect, where one project helps another project, which begets yet another projects — I think that is what sustains enthusiasm and momentum over the long run.
Q: Of all the things you’ve worked on, what do you consider to be the most successful? And can you qualify what you mean by success in this case?
A: Definitely my book, Avogadro Corp. I mentioned that one of my two main ambitions was influencing the world in some way. I’m fundamentally an idea guy, so for me this means getting my ideas exposure, whether that’s implementing them in code or writing about them in fiction.
Since publishing Avogadro, I’ve been blown away by the feedback I’ve gotten.
Brad Feld, cofounder of TechStars, highly praised Avogadro Corp, calling it a “tremendous book”, and recommended it not just in social media, but also in a talk he gave. Amber Case, Portland’s own cyborg anthropologist and cofounder of geoloqi, said a bunch of really nice things about it and the sequel, calling it a “nerd’s dream”. I received an email from a mom one day whose thirteen-year-old son said I was his new favorite author. A colleague read it, and said she had nightmares afterwards. (It’s a technothriller about AI run amok, so this is good.) Ben Huh, CEO of I Can Haz Cheezburger?, endorsed it.
I feel like my ideas have finally reached an audience, and they like them and find them thought-provoking. That’s very, very rewarding.
About William Hertling
William Hertling is the author of Avogadro Corp: The Singularity Is Closer Than It Appears, A.I. Apocalypse, and the forthcoming The Last Firewall.
A fifteen year veteran of the technology industry, he holds ten patents on software and internet technology, developed web and social media strategy at Hewlett-Packard, coached senior executives on social media, and has spoken three times at SXSW Interactive.
An active blogger since 2002, his website, williamhertling.com, receives more than 50,000 visitors a year. He’s been building online communities since 1986 when he ran seven phones lines into the back of his Apple //e to create an online chat system.