IN THE TIME OF ALGORITHMS. A sermon preached by Galen Guengerich All Souls Unitarian Church, New York City January 13, 2013

Transcription

1 IN THE TIME OF ALGORITHMS A sermon preached by Galen Guengerich All Souls Unitarian Church, New York City January 13, 2013 The real winner of the election last November, as it turns out, was not Barack Obama but Nate Silver, the popular forecaster and founder of the political blog FiveThirtyEight.com. Silver, an accountant turned baseball geek turned poker player turned political junkie, correctly predicted the voting for president in all 50 states. In 2008, he correctly predicted the outcome in only 49 states. Predictably, Silver s new book, The Signal and the Noise: Why So Many Predictions Fail But Some Don t, shot up the bestseller lists. While Silver s inbox is now stuffed with lucrative invitations to speak and write, Obama s inbox is stuffed with no-win invitations to solve the unsolvable: the debt ceiling, gun control, Syria, Iran not to mention North Korea, Pakistan, and global warming. Despite Silver s record, he insists that our attempts to predict the future fail more often than they succeed. One reason for our failure is that most of us don t understand probability very well. Silver cites breast cancer screening for women in their forties as a case in point. Three-quarters of women in their forties who have breast cancer will have a positive mammogram, but so will about 10% of women in their forties who do not have breast cancer a 10% false positive rate. So what percentage of women in their forties who test positive for breast cancer actually have breast cancer? The answer is about 10%. How can this be? The actual number of women in their forties who falsely test positive is much larger than the actual number who accurately test positive. You have to factor in the size of the sample and the sensitivity of the test, Silver says. Otherwise, you ll wrongly assume that if you test positive for breast cancer in your forties, you likely have breast cancer. In fact, even if you test positive, the odds are ten-to-one that you don t have cancer. Another reason our predictions fail more often than they succeed, Silver says, is that we have too much information. We can t accurately decide what s relevant and what s not what s signal and what s noise, as he puts it in his title. For example, the US government publishes about 45,000 economic indicators. If you re searching for correlations among these indicators, you have to examine a billion pairs of data. But, he says, the number of meaningful relationships in the data those that speak to causality rather than correlation and testify to how the world really works is orders of magnitude smaller. Silver follows this observation with a statement that stopped me dead in my tracks when I read it: There isn t any more truth in the world than there was before the ~ 1 ~

2 Internet or the printing press. Most of the data is just noise, as most of the universe is filled with empty space. Strictly speaking, Sliver s claim that there isn t any more truth in the world today than there was before the printing press isn t itself true. We know today that the Earth revolves around the Sun, and not vice-versa. We also know that what moves around within the human body is blood pumped by the heart, not humors wafted by musings of a divine spirit. Even so, Silver s insight remains valid. Because we are inundated by data, we easily confuse information for knowledge, facts for truth, the noise for the signal. Several months before Robert F. Kennedy was assassinated in 1968, he gave a speech at the University of Kansas. He spoke about how we evaluate our success as a nation. Our Gross National Product then less than a trillion dollars a year, now more than 15 trillion is one way to judge our national vitality. But Kennedy found the GNP inadequate as a formula for assessment, both because of what it includes and because of what it leaves out. Our Gross National Product, Kennedy said: counts air pollution and cigarette advertising, and ambulances to clear our highways of carnage. It counts special locks for our doors and jails for the people who break them. It counts the destruction of the redwood and the loss of our natural wonder in chaotic sprawl. It counts napalm and counts nuclear warheads and armored cars for the police to fight the riots in our cities. It counts Whitman's rifle and Speck's knife, and the television programs which glorify violence in order to sell toys to our children. Yet the gross national product does not allow for the health of our children, the quality of their education or the joy of their play. It does not include the beauty of our poetry or the strength of our marriages, the intelligence of our public debate or the integrity of our public officials. It measures neither our wit nor our courage, neither our wisdom nor our learning, neither our compassion nor our devotion to our country. It measures everything, in short, except that which makes life worthwhile. And it can tell everything about America except why we are proud that we are Americans. Lest you think my sermon this morning is a screed against statistics, I assure you it is not. I started college as a math major, my wife Holly is a physician, and my daughter Zoe is studying physics. Not only that, our cat Everest often sleeps atop a stack of periodicals that routinely includes Astronomy, Scientific American, and Physics Today. We are a science-friendly household. I will also hand in tomorrow the corrected page proofs of my book God Revised: How Religion Must Evolve in a Scientific Age, due out in late May. Among other pressing issues, the book addresses the question of how we know what we most truly know. The traditional religious answer is clear: our most definitive knowledge comes ~ 2 ~

3 from God as articulated by scripture. The influential fourth-century Christian theologian St. Augustine insisted that, in order to understand the world, we must first believe the Bible. For Augustine, the Bible provided the foundation upon which both reason and experience rested. If the Bible says that Jesus raised Lazarus from the dead, we have to adjust what logic and experience tell us is possible in order to accommodate the more fundamental fact that, according to the Bible, Jesus actually raised Lazarus from the dead. Augustine s view held sway for more than a thousand years. Most people during this period the Middle Ages couldn t even read the Bible, which allowed the church to wield monolithic authority on the question of what was true. This strangle-hold began to weaken during the philosophical awakening we now call the Renaissance and the religious revolution we call the Reformation. Today, as a full-fledged child of the Enlightenment, I firmly believe that reason and experience should trump biblical revelation as our most fundamental source of knowledge. That said, I don t believe the meaning of life lies at the bottom of a spreadsheet, or in the conclusion of a proof, or at the end of an algorithm. As both Kennedy and Silver suggest, we fail ourselves and our aspirations when we confuse measurement with meaning and allow the noise of data to drown out the signal of purpose. One of the most disturbing articles I have read in a long time concerns online dating services. Written by Dan Slater and published in the current issue of Atlantic magazine, the article is titled A Million First Dates: How online romance is threatening monogamy. I had always hoped the opposite would be true of online dating services. If it were easy for single people to meet other single people with whom they might actually be compatible, I thought, the odds of a good match might improve and the incidence of divorce might decline. Alas, Slater found otherwise in his research for a book titled Love in the Time of Algorithms. If you can hope to find an ever-more-compatible mate with the click of a mouse, why not keep looking, and looking, and looking? If the going gets tough in one relationship, why not get going and find another? Most of the online-dating-company executives Slater interviewed agreed that the rise of online dating will mean an overall decrease in commitment. They also admitted that, as Slater puts it, the profit models of many online-dating sites are at crosspurposes with clients who are trying to develop long-term commitments. In practice, this means that profiles of enticing potential matches keep landing in your inbox even after you re matched. A psychologist at eharmony acknowledges that commitment is at odds with technology. You could say online dating allows people to get into relationships, learn things, and ultimately make a better selection, she says. But you could also easily see a world in which online dating leads to people leaving relationships the moment they re not working an overall weakening of commitment. ~ 3 ~

4 The CEO of Match.com s parent company observes that, historically, commitment has been the goal in relationships. Today, that s no longer necessarily true. You could say online dating is simply changing people s ideas about whether commitment itself is a life value, he says. The founder of a dating site based in the U.K. agrees: I often wonder whether matching you up with great people is getting so efficient, and the process so enjoyable, that marriage will become obsolete. Slater frames his article with the story of a man he calls Jacob, who seems to illustrate the well-documented finding that having too many options makes us less satisfied with whatever option we eventually choose. Surrounded by almost-countless romantic prospects, whom he pursues both serially and in parallel, Jacob wonders: When does this end? At what point does using all these relationships to learn more about what I want and how relationships work become an excuse for not putting in the effort to make a relationship last? We fail ourselves and our aspirations when we confuse an algorithm with an intention and allow the noise of data to drown out the signal of purpose. Make no mistake: I m not opposed to online dating as such, any more than I m opposed to calculating economic statistics or assigning political probabilities. These are tools we can use to good effect if they are tools we need to use in order to make decisions we need to make. But algorithms and other procedures we use to solve problems are useful only if we need to solve those problems. If not, they re simply noisy distractions. The reason we gather for worship each Sunday is to shut out the noise, if only for an hour. We ask where we are going and why we are going there. We acknowledge our failures and shortcomings. We focus on intention and ask about purpose. It s true that an online dating service can help you find a potential romantic relationship in a heartbeat. But maybe you don t need a new relationship. Perhaps you need to heed the call to commitment and redouble your efforts to enliven the relationship you re in. It s true that a pollster can tell you whether or not a legislative initiative will win in a particular district or state. But maybe the initiative you need to support doesn t have much of a chance of winning anywhere. Perhaps you need to heed the call of courage and redouble your efforts to support the initiative anyway. It s true that an accountant can tell you whether or not a particular business strategy will improve your corporate bottom line. But maybe you should initiate the strategy even if it hurts your bottom line. Perhaps you need to heed the call of conscience and pursue a strategy that doesn t view the bottom line as the only line that matters. It s true that an online shopping site can tell you where to get the best deal on the latest shirt. But maybe you should patch your old one and wear it instead. Perhaps you should be doing something other than shopping. ~ 4 ~

5 It s true that a GPS can tell you the shortest or fastest route from here to there. But maybe you need to take the long road and meander, or even lose your way. Perhaps it s not worth going at all. In the quiet of this hour of worship, we listen for a signal that reminds us of our purpose. Worship issues an invitation to transcend the noisy confines of our daily lives and view the larger reality of the world and our place in it. Transcendence is an experience of God of feeling deeply connected to everything: all that is present, as well as all that is past and all that is possible. When we hear this signal, we know two things. No matter our past, we are worthy of being loved. And no matter our present, we are brimming with possibility. We have a place of refuge and a source of hope. The chances of this signal being true, by my calculations, are one-hundred percent. ~ 5 ~