Software will eat democracy

It’s easy to be pessimistic about our collective future, but this attitude misses one important trend: software is eating the world.
The form of government which communicates ease, comfort, security, or, in one word, happiness, to the greatest number of persons, and in the greatest degree, is the best.
— John Adams
On a spring afternoon in 2011, a Manhattan public school teacher named Alberto Willmore was arrested in front of his home in the Bronx for possession of marijuana. Willmore, an art teacher, was working on a piece of chalk art when he flicked a cigarette butt in the direction of the nearby sewer grate. A passing NYPD officer saw this and quickly pulled around. The officer restrained Willmore and retrieved a cigarette butt from the area. Willmore was arrested, and lab tests later showed that the cigarette butt retrieved by the officer contained 0.2 grams of marijuana.
This marked the start of an ordeal that would shatter Willmore’s career.
The day after his arrest, he received a letter in his school mailbox informing him that he would be suspended from teaching pending resolution of his case in the courts. At the time, he couldn’t know that his course through the legal system would be an absurdity of delays and extensions lasting nearly two years. It wasn’t until January of 2013, seven court dates and 579 days after the arrest, that Willmore’s case was resolved — with a dismissal.
Finally ready to resume teaching, Willmore was shocked to discover that the city’s Department of Education remained unsatisfied. The Department decided that the nature of the charges against him warranted further suspension pending a hearing. And in March of 2013, the Department made its final ruling. Willmore’s job was gone for good.
By all accounts, Alberto Willmore was an excellent teacher — one that a society bent on maximizing the happiness of its citizens would want to cherish. A student said of Willmore, “I honestly think he would just jump in front of a bullet for us. Like, he loved us, for real.” But instead of cherishing him, Willmore’s government struck him down and dragged him through the gutter.
And Willmore’s case is far from unique. In 2011 there were 759,000 marijuana arrests in the United States. That’s about one arrest every 40 seconds. Each arrest carries with it all of the expense, embarrassment, loss of liberty, and collateral damage to one’s personal and professional circumstances that Willmore experienced.
And here’s the rub. All of this is due to a misguided policy decision. Marijuana prohibition is a bad idea and an empirical failure. Marijuana is now recreationally legal in two states. The sitting president has said that in his opinion marijuana is no more dangerous than alcohol. After an extensive review of the medical research and a series of visits to people affected by marijuana prohibition around the country, CNN’s Sanjay Gupta recently changed his mind on marijuana, saying, “We have been terribly and systematically misled for nearly 70 years in the United States, and I apologize for my own role in that.”
CNN’s Sanjay Gupta recently changed his mind on marijuana, saying, “We have been terribly and systematically misled for nearly 70 years in the United States, and I apologize for my own role in that.”
But we live in a democracy, and prohibition has been our will. Only recently has the country begun to realize that we got this one wrong.
This error is exactly the kind of thing political theorists expect from democratic societies. Policy decisions are hard. The voting citizens in a democracy have jobs and lives to take care of — we don’t have the time, let alone the resources and motivation — to review every policy issue and come to an informed position. So we take shortcuts; we absorb our opinions from the zeitgeist. Or worse, we don’t even bother. We can be lazy and apathetic and are often misinformed and manipulated.
And as a result, we get our preferences wrong. Consider the relationship between a parent and child. A mother doesn’t take it as her job to satisfy all of her child’s preferences — the child often desires the wrong thing. We adults are in the same position. A moment’s historical retrospection is enough to shake free the illusion that we live at an exceptional point in history in which we finally have it all figured out. Smart people thought slavery was right before they realized it was wrong; they opposed universal suffrage before they were for it; the majority of Americans thought invading Iraq was a necessity before we realized we shouldn’t have done it; well-meaning legislators thought alcohol prohibition was good policy before they realized it was not. Larry Page, CEO and co-founder of Google, offers a reflection. “Consider our own history,” he said. “When we started Google, it wasn’t really obvious that what we were doing wouldn’t get regulated away. Remember, at the time, people were arguing that making a copy of a file in a computer’s memory was a violation of copyright. We put the whole web on our servers, so if that were true, bye-bye search engines. The Internet’s been pretty great for society, and I think that 10 or 20 years from now, we’ll look back and say we were a millimeter away from regulating it out of existence.”
For as long as the idea of democracy has been around, theorists have seen the ignorance of the electorate as its inextricable flaw. “The best argument against democracy,” Churchill is rumored to have said, “is a five-minute conversation with the average voter.”
“The best argument against democracy,” Churchill is rumored to have said, “is a five-minute conversation with the average voter.”
But this problem may not be inextricable after all. There’s a revolution going on that Churchill could not have foreseen. It’s been called the digital revolution, the software revolution, and the information revolution. Whatever you want to call it, its impact is unmistakable, and it’s only just begun. The most illuminating summarization of what’s happening may be Marc Andreessen’s famous phrase: software is eating the world. The idea is simple. For any problem within the sphere of human interest that involves the handling of information, from selling a dresser, to finding a spouse, to delivering advertisements to consumers, software presents the opportunity to do things far more efficiently than traditional methods. And whatever software can do better, software inevitably will do better.
The ignorance of the electorate is fundamentally an information problem. As such, it’s a problem that’s ripe for consumption by software. Framed in terms of a solution, it is the problem of discovering the informed preferences of the individual voters. Here’s what I mean by that. As our evolving attitudes towards marijuana prohibition illustrate, our current preferences aren’t the final answer — we may be misinformed, misled, or ignorant. And in that case, our votes will favor policies that differ from those that we’d support if we knew what was good for us. But our informed preferences — those policies that we would favor if we were to take the time to investigate the question thoroughly — are hard to get to. We discover them only by a process of labor-intensive inquiry, which we often don’t have the ability or inclination to undertake. But the days aren’t getting any longer, and the issues aren’t getting any simpler. This is why theorists have long seen democracy’s ignorance problem as inescapable. The challenge for software, then, is to find new ways to streamline the path to our informed preferences. If software can do that, then the electorate’s ignorance will be radically undermined.
This is no easy problem to solve. However, if we hold this problem in mind while taking a careful look at the recent trajectory of progress in information technology and brain sciences, I think we will see that only a very brash person would predict that the path to our informed preferences will remain as inefficient as it has been thus far.
Responses