It’s been quite some time since my last post, which was itself a wrapper around something I’d written in late August, 2012. A lot can happen in half a year.
After the events described in that post, I spent the second half of 2012 wandering. Some of that time was spent with friends and family, but for the most part I was on my own. I was in Iceland, Berlin, New York, St Louis, London, DC, Seattle, Boulder, Chicago, Boston, down and up the waters of the Atlantic. I slept in tatty artist lofts, high-rise condo apartments, hotels dowdy and chic, minuscule bare-bones studios, in the homes and on the fold-out beds of friends, in a nearly-abandoned row house, and even rolling with the waves in a cabin on a cruise ship. A night here, several weeks there, never a home to return to.
At every stop: museums, architecture, oddball excursions, natural and urban wonders. Many meals alone, peering over reading to watch passersby and take in conversations I couldn’t translate. I spent some much-cherished time as a resident at Hacker School, coding with, teaching, and learning from programmers of all ages, backgrounds, and experience levels. In every city I met with entrepreneurs and engineers. I met with them to listen, to brainstorm, to connect, to be convinced or dissuaded, to learn, and to help uncover what I might do next.
Days after the pain and loss of 2012 finally gave way to the clean promise of 2013, I found myself apartment-hunting in the beachside community of Venice, part of the diverse and deliciously surreal sprawl that is Los Angeles. Without work to pick a place for me, a big city with nice weather seemed as good a reason as any to unpack for a while. Then, a day after I’d hung my last picture in my new rental, I got on the phone with Julien Smith at the behest of my friend Evan Prodromou.
Julien has an idea. That idea’s name is now Breather. When Julien described Breather to me, my reaction was one of visceral and unrestrained enthusiasm. When we got off the phone, I paced around my new apartment. “Fuck,” I said, several times to myself and once or twice to a half-dozing cat. “Fuck. This is what I should be doing.”
I am now co-founder and CTO of Breather. I debated myself many times on taking up that role again, having spent much of the past few months thinking that I’d go back to a straight-ahead engineering position in a more established company. On further reflection, I’ve come to accept that early-stage companies are now in my blood. When I feel that a fledgling idea must be brought into the world, it’s simply a matter of how I can best help to make that happen. In Breather’s case, it’s by being CTO, and I feel utterly privileged to be joining Julien and our accomplished Head of Product, Caterina Rizzi, in that capacity.
In time, when we feel we can best introduce Breather to you, we will. It’s a business that lends itself particularly well to being based in New York City, so despite having just gotten settled in LA, I’m packing up once more and relocating to Manhattan by mid-March. I greatly enjoyed my brief time in LA, but the sunshine and California cool seems fairly traded for the dense and unceasing network of people and experiences that is NYC.
I look forward to sharing Breather with you as we bring it to life. If you’re looking for a new challenge – particularly but by no means exclusively in the areas of security, operations, and mobile engineering – get in touch with me: alex@breather.com.
—Mar 07, 2013
The following was written in August, 2012. It was published a month ago in the first issue of Marco Arment’s iOS-only publication The Magazine. Marco has generously allowed his contributors to retain copyright and the option to share their work on their own sites after a time. If you have an iOS device, please consider supporting Marco and The Magazine by purchasing a subscription.
· · ·
I am in a stranger’s apartment in Reykjavik, and for the first time in almost five years, I am truly alone.
Two months ago my life was as per David Byrne’s enthusiastically confused yelps: beautiful house, beautiful wife, not one but two large automobiles. I had a job that seemed like the job I should have at a company I was proud of. I lived in a city that I had carefully chosen to live in, a city that I thought was going to be my home for some time.
Nobody’s life ever really falls apart, exactly. Lives unravel, thread by thread. First, I came to realize that the job wasn’t the right job. Then the city wasn’t the right city. Two threads loose, easily stitched back in; there are other jobs, other cities. Our house went on the market. I resigned.
Then, a month ago, the person I thought I was going to spend the rest of my life with came home from a neighborhood-scouting trip to a place we were intent on moving back to. “We need to talk”. Never good. “I can’t do this anymore. We’re just too different”. The seam, ripped.
That was Saturday evening. There was no going back and no way forward. By Tuesday afternoon we’re in a lawyer’s office, drawing up a settlement. The court moves quickly when there’s no children and no disagreement over who gets what. The divorce should be final by the time this is published. I will probably hear about it by email, wherever I happen to be. I will look at my iPhone, wince, hit “Archive”, close my eyes, and breathe.
The next four weeks disappeared like a long weekend spent sick. I went to the other side of the country while she moved out. Family and friends did their best to hold me together as the stages of grief washed over me in no particular order. The last listed stage of grief in the literature is Acceptance. It didn’t take long to accept that this is all for the best. A gift, wrapped in barbed wire. I just wish I had seen it being wrapped.
Teeth gritted, I went back to the city that was no longer our city, back to the house that was no longer our house. I packed up what was left, put all but necessities into storage. Sold my car. I could have stayed, but I couldn’t stay. The whole city was our life together, brief as it was. Coupled a little over four years total, just a bit over two of which was spent married. We moved mere weeks after the wedding, bought the house not long after. Friends have been engaged longer than our entire relationship lasted. It was something small that we foolishly made into something big, and it’s already receding into the distance. I’m filling my vision with fields of lava, mossy valleys steamed by hot springs. Trying not to look back.
With my life in that place done, I got on a plane, and then another plane, and now I’m alone in an apartment in Reykjavik. Now, this is my life. Not the apartment or the city or travel, but the laptop I’m typing on. We are alone together, again.
· · ·
I owe my life to technology.
I first realized it in my early twenties. Everything important around me at the time, I’d found on Craigslist: my girlfriend, my job, my apartment. It was a powerful realization: I could sit down with my laptop and, in a matter of hours or days, change my world in both superficial and fundamental ways.
That was years ago. Technology specializes over time. The life I just finished packing up wasn’t courtesy of Craigslist. It wouldn’t be, now. The modern web has six sites for everything, branded and polished and localized and full of options. House from Redfin. Cars negotiated online before ever walking into a dealership. Wife from OkCupid. Wedding invitations by email. Date-night dinners booked on OpenTable. Fast and friction-free.
I spent four years telling anyone who asked how we met that OkCupid’s matching algorithms must have been off. “We were only a seventysomething percent match, with like a twelve percent chance of being enemies. Guess they need to work some bugs out!” The joke’s on me, of course. I emailed the right person at OkCupid to apologize for the years of disparagement.
I could blame technology. Maybe stitching together a durable life takes physical work, needle callouses. Maybe technology made it all too easy to slide into a life I wasn’t meant to have. It would be so convenient to think that way. Marriage didn’t work out? Blame the dating site. Bad experience at a restaurant? Blame Yelp.
Technology is just people, though. People like me. We get it wrong. And even when we get it right, people are people: they ignore the algorithms and recommendation engines and scores and weightings. People do what makes sense to them, then and there. What else are we to do? All the specialization, all the options, and we still have to choose. We still have to make our way in the world, alone, save for our technology built from other people’s frozen choices.
· · ·
I will owe the next part of my life to technology, but I will owe it more to experience.
I tried to imagine what my life would be like in the wake of all this if I had been living two hundred years ago. Most likely, I would be trapped. I would be living in the scraps of the life that had unraveled around me. I could not seek the support of friends from around the world at any time of day or night. I could not book passage to wherever I felt I needed to escape to. I couldn’t work from wherever I happen to end up. Trapped.
I am grateful that I live in this time. For all the loose threads in my life, I don’t feel trapped. Technology allows me to make choices that can shape and reshape my world. I don’t yet know what I will do, where I will live, or who I will love. For each of those decisions, though, there are tools, and the tools present options.
The trick, of course, is having the experience to inform the selection of the right options. For that, we have no technological substitute. You simply have to live, work, love, explore, fail at all of those things, and learn.
From Reykjavik to Berlin. Then on to New York City, maybe for a month, maybe for a lifetime. Airbnb, Hipmunk, TripIt. Messages in my inbox suggesting new ventures, new possibilities. Choices. Threads to be sewn together, making something whole again.
· · ·
Postscript, November 2012. A stranger’s apartment in Brooklyn.
I thought long and hard before submitting the above to Marco. Divorce is a subject that, despite its sad prevalence, remains taboo. Divorce is generally discussed behind closed doors, on long phone calls, or via whispers in crowded restaurants. I do not regret my decision to be public about my experience, though. I received many responses, often from people who are going or had gone through something similar. Those people found comfort in knowing that they were not alone, and that is justification enough.
My parents divorced just as I was on the cusp of adolescence. There was nothing in the world that I wanted less than to experience that again, much less firsthand. Some months on, though, I stand by my statement that this was a kind of gift. I do not live with shame, for I have nothing to be ashamed of; I was true and honest, and that was not enough. I do not live in misery. My life today is not at all what I imagined it would be, but it is wonderful and strange and exciting.
It is taking time to make things whole again, as it should. I’m in no rush. I spent so much time trying to organize the life that I thought I wanted. It wasn’t the same as living.
—Nov 11, 2012
Over two years ago, I flew across the country to spend hours in a basement in Brooklyn talking with two guys I barely knew. Those two guys were Josh and Shamir, respectively CEO and CFO of Simple (then BankSimple). When we first sat down, they had already put in months of work figuring out how to start a new kind of banking service. Josh and Shamir were ready to turn their vision of online retail banking with stellar technology, design, and customer service into a reality. I was ready for a new role and a big challenge.
During those first conversations, I asked Josh and Shamir where they saw themselves in ten years. Neither hesitated to reply “running this company”. Since then, I’ve witnessed their perserverance and dedication firsthand. Over the past two years the product has evolved from rough prototypes and a list of features on paper to a working, growing banking service. Today, Simple is bringing on customers by the thousands and rapidly becoming the center of their financial lives. It hasn’t been an easy thing to launch, but working on a product that addresses real problems at the intersection of economics and technology has been a fascinating opportunity.
While I’m extremely proud of the work everyone at Simple has done on the road to launch, I’ve come to recognize that the role of CTO isn’t right for me at this point in my career. Part of my motivation for joining Simple was to explore what it’s like to be a founder and executive at a startup. Being CTO afforded me the opportunity to work with and support a number of the most talented people I know. In that and many other respects it was a great experience. As with any job, though, there are tradeoffs.
The biggest tradeoff for me is that this role has taken me further from learning more about the craft of programming. Simple offers many great career opportunities for engineers, but not in some of the areas I’m most passionate about, particularly programming languages and developer tools. I feel that I have an enormous amount left to learn as a programmer, and hopefully a few things to contribute in that capacity as well. For that reason and some personal ones, I’ve stepped down as CTO at Simple.
My relationship with the team at Simple remains strong. The engineering organization there is in extremely capable hands. I have every confidence that the team will continue to do exceptional, industry-leading work. For my part, I’m staying on as an advisor to the company.
My next steps are to take some time off, relocate, and to consider my next move. This will be the first time in my adult life that I haven’t immediately gone from one job to another, and I’m looking forward to the perspective that a breather will lend.
I’d like to extend my thanks to Josh and Shamir for taking a chance on me, to the exceptional people who came to work with us, and to everyone who’s helped Simple along the way.
—Jul 25, 2012
When I was a kid, I had a morning routine with my family. Over breakfast, we’d divvy up the newspaper. I’d go straight for the Business section, and from there to the back pages with yesterday’s stock prices. Scanning down the lines of tiny text, I’d look for SUN, ORCL, MSFT, AAPL, SGI. Then back to the front of the Business section, reading whatever I possibly could about the silicon giants of the day. Then, finally, the funny pages. I was a weird kid.
Today, like many, I no longer get a physical newspaper. Most mornings I sit down with my iPad and pull up Techmeme, which more often than not leads me to several different “tech blogs” that deliver a mix of rumor, opinion, gossip, rehashed press releases, and, very occasionally, actual news.
There’s a striking difference between the technology news I grew up with and a lot of what I end up reading these days. The difference is something I only recently cemented for myself: most so-called tech news isn’t about technology at all anymore.
There was always something that bothered me about TechCrunch, and it bothered me more and more over the years that the blog rose to be the sort of force that could wipe a site out with a tsunami of traffic just by linking to it. For a publication with “tech” in the name, technology only ever seemed ambiently present in TechCrunch’s reporting. Software and hardware are background music for the site’s real interests: money and power, success and failure, who’s up and who’s down. TechCrunch often read to me like the parts of the Business section I used to skip as a kid, the boring ones with “merger” or “acquisition” in the headline and a picture of two suits shaking hands.
TechCrunch’s popularity urged on a slew of imitators, and now the majority of “technology journalism” consists of blogs that operate in the same general mold. Even newspapers have adopted TechCrunch’s overall format and pace of publication for much of their writing on technology. This has shifted not just the tone and the depth of technology reporting, but the subject matter as well. Where newspapers tended to focus on established technology companies, TechCrunch has always primarily been about startups. With everyone following in their footsteps, startups have become synonymous with technology journalism.
And why not? Startups are exciting. They’re often headed up by interesting people who want to take on big problems. Startups seem somehow rebellious and alternative, but in a way that’s safe and accessible enough to read about at the breakfast table. Startups have drama. Will they survive? Get bought for a billion dollars? Fail spectacularly? What’s more, startups love to talk about themselves, which makes life easier for reporters. What founder, save one who’s already well-established and cagey, is going to turn down free publicity?
Startups, however, are not inherently technology companies.
It’s now accepted-going-on-cliché to say things like “software is eating the world”, which is an aggressive way of assuming that every company now has to be at least a bit of a technology company, and those that want to grow rapidly even more so. Many new companies targeting industries as diverse as eyeglasses and baby food are, at the outset, leveraging technology for everything they do: supply chain management, marketing, recruiting, internal communication, product development, and so on. This makes these businesses look like technology companies, if you squint. But, of course, they aren’t. They’re eyeglasses and baby food companies.
Somewhere along the way, we confused “startup” for “tech startup” and “company” for “technology company”. Now that every growing business requires significant competence in technology to succeed, the distinction is even blurrier. Is a company that has staff members with “programmer” or “engineer” in their titles a technology company? Are they a technology company if they were funded by venture capitalists who have previously funded businesses that we think of as technology companies? Are they a technology company if their founder was using a laptop when she came up with the idea for the business?
These questions are pretty easily resolved. You are a technology company if you are in the business of selling technology. That is to say, if your product – the thing you make money by selling – consists of applied scientific knowledge that solves concrete problems and enables other endeavors, you are a technology company.
By this definition, most of the companies that dominate the “tech blogs” are not technology companies. They’re just, well, companies. These businesses might use technology, or develop technology, or even be run by people who used to work at technology companies, but they don’t exist to create and sell technology.
IBM is a technology company. Basho is a technology company. Boundary is a technology company. Apple is a technology company thanks to some of its lines of business (hardware, software), but is not a technology company in all of its lines of business (music, movies, books). Ditto Google and Amazon, which make money both from selling technology and from leveraging technology to sell things like advertisements and socks.
Simple is not a technology company. We use and develop technology, but we are a banking service, and we make money through banking. Zappos, likewise, is not a technology company. They use and develop technology, but they make money when people buy shoes from them. Technology might make our respective businesses run more effectively, or give us a competitive advantage, but it is not our product. An interest in technology might shape our corporate cultures, but technology is not what we sell.
Put more crudely: sticking feathers up your butt does not make you a chicken, and having an engineer or a data scientist on staff does not make you a technology company. Having people in these kinds of roles just makes you a company that has to do business in a world that requires code to be written in order to operate efficiently.
This definition also doesn’t suggest that only “enterprise” technology companies are “real” technology companies. There are plenty of companies that make money by selling technology to consumers. But there are also a lot of businesses selling actual things to consumers who get lumped in with technology businesses simply because they sell products via the web. That might have been an important distinction in 1995, but no longer.
What I’ve spelled out is a definition, but not a value judgment. I’m not saying that technology companies are good and other types of companies are bad, or vice versa. What I’m saying is that they are different, and that they require different approaches for investors, executives, and particularly for the press. Because technology companies are trafficking in applied science, an understanding of that science is required in order to reason about and report on their operation with any kind of cogency.
“Tech company” and “tech startup” are over-applied labels that have outlived their usefulness. Calling practically all growing contemporary businesses “technology companies” is about as useful as calling the enterprises of the industrial era “factory companies”; it accurately describes an aspect of what they are (or were), but it doesn’t really capture the totality of their operation. It certainly doesn’t tell you anything substantive about how they’ll behave in the market over the long term, which is probably the most useful reason to label a business at all.
What’s more, this mis-labeling results in the conflation of companies in totally different industries applying totally different business models, all being funded and staffed and reported on by the same pool of people. If we remove the label of “tech startup” – and with it the hypothetically stellar trajectory we like to imagine such businesses are on – we’re forced to confront the reality of a business’s model, independent of the reverberations of the echo chamber.
That said, what TechCrunch and other sites often cover is clearly something distinct: companies that have information technology so close to their core (“in their DNA”, if you can still stomach that phrase) that they seem to have acquired the essence of this thing that they use to get ahead. But in confusing how these businesses make their product and what their product is, we introduce a whole set of assumptions and biases and information gaps that, by the looks of it, have resulted in a distorted market for companies and their equity, salaries, publicity, office space, and so forth.
I’m not sure what the ideal terminology is for today’s burgeoning companies. At startup showcases, terms like “disruptive businesses” are thrown around, but they presume too much to be useful. The word “startup” itself presumes too little, and has become overloaded. In time, a new word will present itself. Or maybe we’ll have to learn to be satisfied by talking about businesses that leverage information technology in creative ways as, simply, “businesses”. But clarifying our language is only the outward expression of clarifying our thinking.
Really, nobody but sociologists and historians should be running around talking about “technology” in the large. It’s a vague word for an even vaguer set of ideas about how humans operate in the world. Our current culture sprinkles this word around like faerie dust, blinding us to the truth of what things are, how they work, and who’s responsible for them. Broad cultural change takes time. As a starting point, the business world seems as good a place as any to wipe away the magic dust and start seeing clearly.
—May 08, 2012
A few years back, I lived and worked in Northern Virginia, a sprawling collection of suburbs just outside Washington, DC. It’s a place of malls and strip malls, office buildings and office parks, a territory inlaid with and surrounded by perpetually packed freeways carrying government workers and contractors home to planned communities. It’s a place people live because of a job.
At the time, I had a job I liked very much, and so did some people who worked across the highway-wide street. We all got to do challenging things with computers and we had a lot of freedom as to how we went about it. We became friends.
Some evenings, after work, we’d gather at one or another of the area’s strip mall Middle Eastern cafés. We’d order big plates of kabobs and rice and hummus and pita and try to best one another at the ideal combination of flavors of shisha to share in the hookahs that are the cafés real draw. On warm nights we’d sit outside and watch the sons and daughters of wealthy diplomats preen for one another in an endless contest for the nicest car, the chicest clothes, the most beautiful partner on your arm. Farsi music videos blared and the suburbs around us were all but gone in the hanging smoke.
And then, when the plates of food were replaced with mint tea and Turkish coffee, we played a game.
The game is called Zendo. It’s made by a company called Looney Labs who, like many clever small game companies, reuse a set of universal pieces for a number of their products. Zendo is played with signature plastic triangles called Icehouse pieces, as big and as varied a set as you care. The triangles are hollow, allowing them to be stacked. They have a small “pip” on each side that designates a point value. The triangles come in just three sizes, but are available in a whole variety of colors and opacities.
A round of Zendo begins by designating a “master”. In her mind, the master comes up with a rule that describes multiple configurations or structures made up of the triangular pieces. A rule might be: “blue triangles must always have a red triangle stacked atop them”. Or: “the total number of pips in the structure must be divisible by two”. The master builds one structure that satisfies the rule alongside one structure that does not, and the two structures are marked accordingly with white and black stones respectively.
Each turn, a player who is not the master builds a structure from the shared pile of spare triangles. Each structure a player builds is marked with a stone designating its adherence to the master’s unspoken rule, just as with the master’s original example structures. If a structure fits the master’s rule, an additional green guessing stone is earned by the player that built it. A player can trade a guessing stone for a guess at the rule, asked of the master in front of the other players. Play proceeds until someone correctly guesses the master’s rule.
That, modulo some details, is the game.
Zendo is a game of inductive logic. Anyone can enjoy it, but programmers take particular delight in it. In broad strokes, a round of Zendo mirrors the intellectual process programmers engage in when debugging. You know there’s a rule in the machine undermining the task you’re trying to accomplish, but you don’t know what the rule is, so you build and rebuild and think and observe and rebuild again until you understand the rule, and then you win.
Zendo’s process of becoming gradually un-blind could even be said to be a metaphor for how many of us learn to program, particularly if you’re self-taught. You’re thrown into a black nexus of thousands of unknown rules and made to reason in the dark. With each rule you discover, a little less dark. (I’ve been programming in earnest for sixteen years and presently find myself in a sort of early-evening twilight, plus a couple of sconces in the corner and maybe a night light.)
For me, the interesting thing about Zendo is not its parallels to the task of programming as an individual, but thinking of the game as a model for collective problem-solving.
One strategy for Zendo is to hole up in your mind, obsess over the structures you’ve built, and try to crunch through every possible permutation on your own. This approach works well if you’re a genius, I’d imagine, but most of us are not. It’s also a good way to lead yourself astray, convinced that you can see the general shape of the master’s rule when in fact you’re thinking along entirely the wrong lines.
A more broadly applicable strategy is to throughly observe the plays of your opponents and treat “solving” the round as a kind of computation distributed across all the players. Together, you and your fellow players are brute forcing the problem of discovering the rule. Every (motivated) player at the table decreases the time it takes to win.
The best rounds of Zendo I’ve ever played are ones in which the master has chosen a devilishly hard rule, a rule so difficult to guess that the players cease to be in competition with one another and are instead in a collective war against the inscrutability of the master. When one player finally wins, it’s due to the combined efforts of all the players building structures, making guesses, exploring the solution space. That shared victory is a wonderful moment.
Flashes of brilliance are few and far between. Even when they strike, turning a brilliant insight into something reproducible is almost always an effort beyond just one person. Programming lore is full of stories of lone-hacker-in-the-night heroism, but the reality of what we do is one built on collaboration: between pairs, amongst teams, amongst companies in the marketplace, amongst universities and laboratories, amongst contributors to open source projects.
I write this by way of encouraging you to:
One last thing about Zendo: it’s not much fun to be the master until someone guesses the rule.
—Apr 09, 2012