Complexity Theory: Be careful what you build for

Opinion by Jason Zhao
April 21, 2020, 11:18 p.m.

Part of “Complexity Theory,” a column on the tangled questions of our technological age.

Another beautifully optimized morning: outfit laid out, floors heated to just the right degree of lukewarm and Spotify’s “Daily Mix” dishing out tunes you didn’t even know you wanted to hear. You want to give Siri a hug for all that it does, but that’s a feature in iOS231, and you still haven’t upgraded. Whatever. You ask Alexa to remind you to update Siri this evening. As you leave the house, Google Maps informs you that supersonic aircraft is 34 milliseconds faster than the Hyperloop. You arrive to work on the microsecond, ready to build the future.

In a world where young and old alike are dying from a shortage of masks and ventilators, who wouldn’t prefer such a utopia? As Marc Andreessen rightly points out in his recent essay “It’s Time to Build,” the COVID-19 catastrophe has underscored systemic issues in American society. These issues, I agree, “run deeper than [our] favorite political opponent.” In education, healthcare and transportation, problems manifest obnoxiously in the vast amount of “things we urgently need but don’t have.”

Why do we not have these things? According to Andreessen, it is because we “chose not to build.” As an antidote to our general malaise, Andreessen urges us to emerge from our “smug complacency” and engage in a “full-throated, unapologetic, uncompromised” building spree. If we want a better future, we have to build it first — no excuses.

I am not here to tell Marc Andreessen that I do not also wish for better education, better healthcare and a better society. Lives and livelihoods, no doubt, would benefit from engineers, investors and political leaders collaborating to build the answers to our many problems. I am not here to tell you to stop building.

What I am here to do is slap a large, glaring asterisk onto a well-intentioned polemic that is equal parts inspirational and obtuse. Building is not always good. That much is obvious: From atomic bombs to the Juul, technology has wreaked havoc in ways small and large, intentional and accidental. With the recent “techlash” rippling through Silicon Valley, I am sure that none of us need more dystopian diatribes against the perils of technology.

More importantly, and perhaps less obviously, building is not always the answer. Like all pep talks, Andreessen’s message attempts to put a rosy filter on some ugly truths. By advancing the unambiguous position that “there is only one way … to create the future we want for our own children and grandchildren, and that’s to build,” we risk glorifying solutions before thinking critically about what we’re building for in the first place. The most obvious way a singular focus on building can go wrong is what I’ll call the “Tinder for Uber” phenomenon.

The “Tinder for Uber” phenomenon occurs when people build, for lack of a better phrase, useless crap. Andreessen wants everyone to be “building directly, or helping other people to build, or teaching other people to build, or taking care of people who are building.” While noble, the ideal that we should all be building tends to lose its luster upon the realization that many intelligent people with valuable skills are wasting their time building dumb things like smart umbrellas. Andreessen claims that “capitalism is how we take care of people we don’t know,” which only underscores the problem: capitalism is too often the privileged believing that they’re taking care of people they don’t know — namely, everyone else. Before we start building blindly, we should think more about what — and for whom — society incentivizes us to build.

Building is not an intrinsic good, even if Silicon Valley wants to enshrine it as one. Juicero built a $400 juicing appliance and raised $118 million in venture funding to boot, only for the public to discover that its high-tech juice packs could be squeezed with bare hands: no fancy appliance needed. Less amusingly, Theranos garnered a valuation of $9 billion for its nonexistent blood testing technology before employees were discovered running backhanded tests through third-party machines. Building for its own sake provides ample cover for arrogance to masquerade as altruism. Let’s talk less about building more, and more about building well. 

Not that there is never a time for more; sometimes building more goes hand in hand with building well. With a legitimate manufacturing crisis in our country, now is certainly one of those times. However, as long as builders are more concerned with creating a convenient future than a conscientious one, we will inevitably suffer shortages when the next disaster strikes. Andreessen knows this, and he acknowledges it too. As a prominent venture capitalist, Andreessen likely witnesses the “Tinder for Uber” phenomenon more often than anyone. Indeed, he invites his critics to “conceive [their] own” ideas of what to build instead of criticizing his ideas of what to build. 

However, in asking his critics “what do you think we should build?” and not “what do you think we need?” Andreessen presupposes the solution before understanding the problem. The hallmark mistake of technocratic overoptimism is asking all about where, what, when and how to build without ever stopping to wonder why

Andreessen may understand the “Tinder for Uber” phenomenon, but as his statements make clear, he completely misses the alarming cultural forces behind it. By conflating what we build with what we need, Andreessen falls victim to the “hackathon” fallacy. The “Tinder for Uber” phenomenon is just a particular manifestation of this belief, which portrays our world as a blown-up, society-wide hackathon. In a typical hackathon, participants are urged to come up with technical solutions to a variety of problems under time pressure ranging from hours to days. Hackathons instill a mindset that there is some “hack” to every social problem. 

The hackathon approach works splendidly for corporate software, consumer applications and even technical contributions that chip away at complex issues like climate change. What happens, however, if our problems can’t be solved by building? What should we do when our problems require empathy, humility and compromise instead of one-to-one tutoring, supersonic aircraft and JavaScript? How can we solve racism and economic inequality by building

By putting all faith in building things, the hackathon fallacy admits a corresponding lack of faith in people. It diminishes the vital role of citizens and the political institutions governing them. Apparently too busy yelling at America for its failure to build, Andreessen conveniently glosses over Hong Kong and Singapore’s quiet success against coronavirus. These governments did not contain the virus just by exhorting their citizens to build: Hong Kong’s production has been ravaged by democratic protests, and Singapore’s imports are 150% of its GDP. Instead, leaders mobilized immediately, strong public institutions delivered necessary supplies and citizens — even the ones in Hong Kong who had been protesting for months on end — heeded the warnings of medical experts. Effective political responses enabled these states to succeed where America, with its world-class researchers, hospitals and technology, did not. 

What Andreessen doesn’t adequately account for is that we live in a democracy governed by people, not a corporation governed by a CEO. Democracies cannot, as Andreessen suggests, simply do away with “old,” “entrenched” and “irrelevant” public institutions in an effort to thrust our society into a brave new world. The nature of political institutions is to protect as much as it is to provide, and the Silicon Valley motto of moving fast and breaking things doesn’t work so well when human wellbeing is at stake. People should not be reduced to red numbers on a spreadsheet, profit and lives hanging on opposite sides of an unfeeling equal sign.

Not all problems can be built out of existence by sheer force of will, and crucially, many challenging problems don’t require anything to be built at all. Andreessen recognizes the atrocity of medical personnel resorting to rain ponchos as medical gowns: “Rain ponchos! In 2020! In America!” Yet does he also recognize the atrocity of senseless coronavirus deaths as a result of political corruption? (Fake news! In 2020! In America!) No amount of building is going to solve a lack of integrity, but our votes, our public dialogue and our democracy can make a dent this November.

The problem isn’t as simple as just lacking a desire to build. Portraying San Francisco’s housing crisis as a matter of building more housing ignores the role that wealthy corporations play in gentrification. Depicting education as a matter of “matching every young learner with an older tutor” ignores the fact that deep pockets like Andreessen’s historically balk at paying the higher taxes that might provide even one teacher for every 10 students. Systemic social issues are almost always a problem of negotiating competing values and not merely a problem of building. 

In a democracy, we can’t just build over these value tensions as if they never existed. We also can’t just “force the incumbents to build,” “even if they don’t like it.” Because not all incumbents are speed bumps; some are elected representatives. And perhaps some — God forbid — care more racial justice, abortion reform and foreign policy than they care about the Hyperloop.

It’s convenient to blame Americans for not building enough. It’s much harder — and less provocative — to appeal to the democratic institutions that have been the one truly enduring American innovation. We need builders, surely. But more importantly, we need engaged citizens. Our nation and civilization were not “built on production, on building.” Our Constitution begins not with a copyright, but with inalienable rights. In 2011, Andreessen sagely proclaimed that “software is eating the world.” Almost a decade later, it would be wise not to let it eat our democracy too.

Many thanks to Nik Marda, Troy Shen, Roshan Fernando and Arjun Ramani for reading early drafts and offering thoughtful suggestions.

Contact Jason Zhao at jzhao23 ‘at’ stanford.edu.

The Daily is committed to publishing a diversity of op-eds and letters to the editor. We’d love to hear your thoughts. Email letters to the editor to [email protected] and op-ed submissions to [email protected]Follow The Daily on Facebook, Twitter and Instagram.



Login or create an account