When I get home, I really look forward to watching the string of amusing 3-D videos and fun games that Albert lines up for me. That’s the name I gave to the avatar from Universal who filters my media for me. Albert always gets the coolest stuff because I’ve trained him really well. Ever since high school I would spend at least 10 minutes every day correcting his selections and adding obscure influences, really tuning the filters, so that by now, with all the new AI algos and the friends of friends of friends’ scores, I have the most amazing channel. I have a lot of people who follow my Albert daily. I am at the top of the leaderboard for the VR worlds filter. My mix is so popular that I’m earning some money from Universal—well, at least enough to pay for all my subscriptions.
• • •
We are still at the early stages in how and what we filter. These powerful computational technologies can be—and will be—applied to the internet of everything. The most trivial product or service could be personalized if we wanted it (but many times we won’t). In the next 30 years the entire cloud will be filtered, elevating the degree of personalization.
Yet every filter throws something good away. Filtering is a type of censoring, and vice versa. Governments can implement nationwide filters to remove unwanted political ideas and restrict speech. Like Facebook or Google, they usually don’t disclose what they are filtering. Unlike social media, citizens don’t have an alternative government to switch to. But even in benign filtering, by design we see only a tiny fraction of all there is to see. This is the curse of the postscarcity world: We can connect to only a thin thread of all there is. Each day maker-friendly technologies such as 3-D printing, phone-based apps, and cloud services widen the sky of possibilities another few degrees. So each day wider filters are needed to access this abundance at human scale. There is no retreat from more filtering. The inadequacies of a filter cannot be remedied by eliminating filters. The inadequacies of a filter can be remedied only by applying countervailing filters upon it.
From the human point of view, a filter focuses content. But seen in reverse, from the content point of view, a filter focuses human attention. The more content expands, the more focused that attention needs to become. Way back in 1971 Herbert Simon, a Nobel Prize–winning social scientist, observed, “In an information-rich world, the wealth of information means a dearth of something else: a scarcity of whatever it is that information consumes. What information consumes is rather obvious: it consumes the attention of its recipients. Hence a wealth of information creates a poverty of attention.” Simon’s insight is often reduced to “In a world of abundance, the only scarcity is human attention.”
Our attention is the only valuable resource we personally produce without training. It is in short supply and everyone wants some of it. You can stop sleeping altogether and you will still have only 24 hours per day of potential attention. Absolutely nothing—no money or technology—will ever increase that amount. The maximum potential attention is therefore fixed. Its production is inherently limited while everything else is becoming abundant. Since it is the last scarcity, wherever attention flows, money will follow.
Yet for being so precious, our attention is relatively inexpensive. It is cheap, in part, because we have to give it away each day. We can’t save it up or hoard it. We have to spend it second by second, in real time.
In the United States, TV still captures most of our attention, followed by radio, and then the internet. These three take the majority of our attention, while the others—books, newspapers, magazines, music, home video, games—consume only slivers of the total pie.
But not all attention is equal. In the advertising business, quantity of attention is often reflected in a metric called CPM, or cost per thousand (M is Latin for “thousand”). That’s a thousand views, or a thousand readers or listeners. The estimated average CPM of various media platforms ranges widely. Cheap outdoor billboards average $3.50, TV is $7, magazines earn $14, and newspapers $32.50.
There’s another way to calculate how much our attention is worth. We can tally up the total annual revenue earned by each of the major media industries, and the total amount of time spent on each media, and then calculate how much revenue each hour of attention generates in dollars per hour. The answer surprised me.
First, it is a low number. The ratio of dollars earned by the industry per hour of attention spent by consumers shows that attention is not worth very much to media businesses. While half a trillion hours are devoted to TV annually (just in the U.S.), it generates for its content owners, on average, only 20 cents per hour. If you were being paid to watch TV at this rate, you would be earning a third-world hourly wage. Television watching is coolie labor. Newspapers occupy a smaller slice of our attention, but generate more revenue per hour spent with them—about 93 cents per hour. The internet, remarkably, is relatively more expensive, increasing its quality of attention each year, garnering on average $3.60 per hour of attention.
A lousy 20 cents per hour of attention that we watchers “earn” for TV companies, or even a dollar an hour for upscale newspapers, reflects the worth of what I call “commodity attention.” The kind of attention we pay to entertainment commodities that are easily duplicated, easily transmitted, nearly ubiquitous, and always on is not worth much. When we inspect how much we have to pay to purchase commodity content—all the content that can easily be copied—such as books, movies, music, news, etc.—the rates are higher, but still don’t reflect the fact that our attention is the last scarcity. Take a book, for instance. The average hardcover book takes 4.3 hours to read and $23 to buy. Therefore the average consumer cost for that reading duration is $5.34 per hour. A music CD is, on average, listened to dozens of times over its lifetime, so its retail price is divided by its total listening time to arrive at its hourly rate. A two-hour movie in a theater is seen only once, so its per hour rate is half the ticket price. These rates can be thought of as mirroring how much we, as the audience, value our attention.
In 1995 I calculated the average hourly costs for various media platforms, including music, books, newspapers, and movies. There was some variation between media, but the price stayed within the same order of magnitude, converging on a mean of $2.00 per hour. In 1995 we tended to pay, on average, two bucks per hour for media use.
Fifteen years later, in 2010, and then again in 2015, I recalculated the values for a similar set of media using the same method. When I adjusted for inflation and translated into 2015 dollars, the average cost to consume one hour of media in 1995, 2010, and 2015 is respectively $3.08, $2.69, and $3.37. That means that the value of our attention has been remarkably stable over 20 years. It seems we have some intuitive sense of what a media experience “should” cost, and we don’t stray much from that. It also means that companies making money from our attention (such as many high-profile tech companies) are earning only an average of $3 per hour of attention—if they include high-quality content.