7.5. Information Overload
As Calvin Mooers told us back in 1959, people may not want information, because having it can be painful and troublesome. When it comes to information, sometimes less is more, as we see in Figure 7-3. We know this explicitly from studies that show an inverted-U relationship between the volume of information and decision quality. In fact, a recent study at Kings College in London showed that information overload harms concentration more than marijuana.[*] We also know this tacitly from experience. We have all felt overwhelmed by details, and we all choose, every day of our lives, to ignore vast quantities of data.
Figure 7-3. The inverted U
We choose not to choose. We rely on habit. We trust familiar brands. We copy our colleagues. But the decisions only multiply fastereducation, entertainment, insurance, investmentwe are inundated with products, services, plans, and promotions. And it makes us miserable. As Barry Schwartz explains in The Paradox of Choice:
7.6. Graffiti Theory
We're starting to understand the pathology of information consumption and its long-term effects on decisions, thanks to novel insights flowing once again from AI. Our guide to this part of the maze is Jeff Hawkins, founder of Palm Computing and Handspring, architect of the Palm Pilot and Treo, and inventor of an alphabet called Graffiti, shown in Figure 7-5.
Figure 7-5. Graffiti by Jeff Hawkins
He has made great contributions to mobile computing, yet our interest lies not in Jeff's past but in his passion, which he confesses in his book On Intelligence:
Jeff argues that "AI suffers from a fundamental flaw in that it fails to adequately address what intelligence is or what it means to understand something." His quest for understanding has led him, through the fields of computer science, biophysics, linguistics, and neurophysiology, to an intense focus on the neocortex:
And Jeff's path-breaking work at the Redwood Neuroscience Institute has led to a new model of intelligence, the memory-prediction framework:
The neocortex stores hierarchical sequences of patterns in invariant form, and recalls those patterns auto-associatively. This lets us recall complete patterns when given only partial or distorted inputs. In the reflection of a flawed ring, we see Plato's perfect circle. In the myriad breeds of canine familiaris, we recognize the category of dog and the tree of Animalia, Chordata, Mammalia. And in three notes, we find the words to a song.
Of course, memory is only half the story. Its conclusion lies in the future:
Jeff argues convincingly that when we take a step, catch a ball, read a poem, or write a book, we draw upon experience to make predictions. We use the past to see the future. Input begets output. Information shapes behavior. Which brings our train of thought back to graffiti, in the form of illegal art on the streets and subways of New York City, shown in Figure 7-6.
Figure 7-6. New York City subway graffiti (The graffiti epidemic in full swing. September 1980. Photo by Steve Zabel, collection of Joe Testagrose. From http://www.nycsubway.org/)
In the Tipping Point, Malcolm Gladwell explains the remarkable drop in crime rates in the Big Apple during the 1990sin five years, total crime fell by 50%, murder by 65%as evidence that little things can make a big difference. Specifically, he makes the case for the power of environmental context under the rubric of "broken windows" theory:
In other words, exposure to seemingly insignificant misdemeanors over time can forge a psyche more prone to violent crime. Gladwell uses the 1984 shooting of four unarmed teenagers on the NYC subway by Bernhard Goetz as a case in point. Goetz's split second decision to shoot in "self-defense" resulted from years in an atmosphere of pervasive lawlessness. As his biographer notes, the bullets were "aimed at targets that existed as much in his past as in the present." Input begets output. Information shapes behavior.
Which brings us to graffiti theory, my corollary to both memory-prediction and broken windows, which suggests that all information that flows through our senses continuously and unconsciously shapes our memories, beliefs, predictions, decisions, and behaviors. We are born with instinct, but in matters of intuition, we are lifetime learners. Information is data that makes a difference, literally. It changes our minds, physically.
This is why I feel a twist in my gut when I think about today's standard information fare. At a time when the old-line media of newspapers, radio, and TV broadcast outlets are controlled by a handful of corporations, there's enormous potential for folded feedback, as the predictions of a few become input for the many. We know this from research:
And, we're seeing it already in the growing polarization of beliefs, as we become victims of repetition and rhetoric. Politics shifts gears from differing opinions to different matters of fact. Medicine gets co-opted by drug sellers and spin doctors. Evolution is just a theory, like global warming and weapons of mass destruction. It's hard to find the truth, especially in a Super Size Me culture slap happy on soft think and content nuggets.
But if you look real hard, you can find the truth, or at least that's the theory driving Steven Levitt, one of the world's most brilliant and unusual economists. Levitt's bizarre curiosity, along with a gift for regression analysis, makes him a perfect candidate for exposing the untruths of conventional wisdom. For example, he really turned heads in 2003 by proving that "If you both own a gun and have a swimming pool in the backyard, the swimming pool is about 100 times more likely to kill a child than the gun is."[*]
Levitt's real passion is crime, and his search for answers leads us back to the "broken windows" of New York City. He studied the same crime drop as Malcolm Gladwell, but Levitt's investigation went deeper and wider. Crime rates did indeed fall precipitously in the 1990s, but the trend began before the police cleaned up New York City, and similar drops occurred simultaneously in cities throughout the United States. Nationwide, nonviolent crime fell by 40% and the teenage murder rate fell by 50%.
So, it's unlikely that policing strategies in New York City deserve much credit. And, one by one, Levitt proves most of the other popular explanations to be false as well.
So, how does he explain it? In a word, abortion. Levitt argues that the U.S. Supreme Court's 1973 ruling in Roe v. Wade, which legalized abortion, resulted in an immediate decrease in unwanted children, which led to the drop in teenage criminal activity 15 years later. Steven Levitt makes a powerful case. It's certainly hard to dispute his data. And yet, his truth is an uncomfortable one. This is the type of information Calvin Mooers knew people may not want. This is the kind of news that the many may choose to ignore.
The power of our culture and our surrounding information environment to mold us is nothing new. As Herbert Simon noted:
But what is new is the level of control we each have over our sources of input. And the Web, in particular, provides access to myriad sources of news, opinion, and data from all over the world. Google News delivers Aljazeera.net, Blogcritics.org, and CNN.com. A standard keyword search opens the widest of windows on the narrowest of topics. We have the power to inform ourselves like never before. We can manage our information diet, and thus the health and well-being of our rational and intuitive decisions.
So, when it comes to the problem of uninformed decisions, I'd like to say the Web is the solution, but that would be only a half truth, for it can just as easily perpetuate ignorance. In addition to a bias for free and digital, power laws and preferential attachment create dominant hubs and fragmented discussion. A few companies and people capture most of the eyeballs. And in this realm of high precision and low recall, only 15% of web pages include links to opposing viewpoints. Internet topology features countless islands of discourse, isolated by social structure and semantics. For those without a compass, these islands become breeding grounds for misinformation and apophenia.
It's at these degrees of confluence, where information feeds ignorance, that you'll find librarians ranting about literacy. As experts in the critical selection, evaluation, and use of information, we're worried by people's preference for the Web over the library. We're concerned by users who ignore the influence of advertising on source reliability. Why search the Web when weve got LexisNexis? Why read blogs when we've got books? Why quote Wikipedia when we've got Encyclopædia Britannica? Why choose Google's graffiti ghetto over our scholarly society? Librarians understand what's at stake, and we fear where we're headed. We're canaries in the coal mine, singing a familiar refrain:
Some heed our warning. For instance, a United Nations declaration puts:
And, to some extent, information literacy is being incorporated into K-12 education, where it can make the greatest impact. But today, in truth, most people don't listen. Librarians are like doctors, preaching what's good for us, while breaking their own rules. They use Google profusely and rarely speak Boolean. And, in any case, they're not speaking with one voice. In fact, the Internet only makes the fault lines more pronounced.
I, for instance, lean toward the liberal side of librarianship. I see bias in the facts of the Wall Street Journal and the Encyclopædia Britannica, and I find real value in the opinion of blogs and the collaboratively authored free content of the Wikipedia. Like relevance, authority is subjective and ascribed by the viewer. Thus, I believe our society needs better information literacy, but I'm not sure we have yet built the understanding or consensus requisite to teach. And, I'm sure that when push comes to shove, access trumps literacy. Information that's hard to find will remain information that's hardly found.
In this sense, it's ironic that in the process of writing this book, my first visit to a physical library was to find the last article, "Rational Choice and the Structure of the Environment" by Herbert Simon. For most of my research, I found what I needed from where I sit, via the free Web, online databases, and my personal bookshelf. But this ancient text wasn't available online, not for free or for fee. So, I was forced to visit the University of Michigan Graduate Library in search of Psychological Review, Vol. 63, No. 2, 1956.
This arduous task took a couple of hours, and it wasn't easy, though the thin white line shown in Figure 7-7 was a helpful guide as I negotiated the maze of books to BF1.P7 in the South Stacks on the eighth floor, where I promptly snapped up the article with the camera in my Treo.
And, as I liberated the words of Herbert Simon from their dusty bindings, it occurred to me that the next hands to touch this crumbling text would be working at the behest of Google to digitize all seven million volumes in this canonical library collection.
And, I found myself, once again, inspired by the ambition of Larry Page and Sergey Brin to organize the world's information and make it universally accessible and useful, for these are not just words, but ideas linked to actions with profound social impact.
Figure 7-7. The University of Michigan Graduate Library
I can't imagine how anyone who cares about learning and literacy could not be excited by the goals of Google's Library Project, which are summed up as follows:
The collections of the University of Michigan, Harvard University, Stanford University, the New York Public Library, and Oxford University will be accessible to anyone, any time, anywhere. This is amazing. The world's greatest works of art, history, science, engineering, law, and literature are about to join the conversation we call the Web. Through the lens of graffiti theory, this is a major upgrade to our collective memory, which will ultimately shape our predictions, decisions, and actions. In the end, computers are not about the creation of artificial minds, but the augmentation of real intelligence.
And, our networks are not just about cold hard facts and frictionless data streams, for it's the inspiration in information that separates man from machine. This is why I love the Web. It opens a door to a garden of forking paths, a labyrinth of symbols, wild with ideas and memories and myriad futures present. The Web lets us find our own way. We choose our links and our leaders. We decide where to go, what to believe, and who to follow. Our garden is a maze of heroes and memes, where what we find shapes who we become. Sadly, our story nears its end, but before we part, let's wander through a few more sites and sources that reflect the pregnant promise of ambient findability.