by Paul Allen
But there was a catch, the analyst noted: “The challenge facing cable today is to learn how to effectively market its basic and advanced services in an increasingly competitive marketplace.” As things turned out, the competition was even more intense than anticipated. And we failed to rise to the challenge.
JERRY KENT’S SUCCESSOR, Carl Vogel, worked hard to renegotiate our loan maturities and keep us afloat. The problem was that Charter was equally vulnerable on the operational front. A cable business must continually sweat all the details of customer service, from pricing tiers to just how many technician visits (or “truck rolls”) it takes to get someone’s service issues resolved. From where I sat, Carl wasn’t up to that part of the job.
For years I’d been urging the cable industry to trump our satellite and telecom competition with faster deployment of digital video recorder set-top boxes and video on demand. I knew that consumers would want these features. I had them for years in my home-brewed multimedia system, and my visitors—several industry leaders among them—responded with enthusiasm. But as Cox and Cablevision began to roll out DVRs more aggressively (along with high-speed phone service, “triple-play” TV/phone/Internet packages, and, later, high-definition TV), Charter lacked the funding and focus to keep pace.
Smelling weakness, satellite providers targeted our areas with promotions and extra advertising. We lost subscribers by the bucketful, half a million during Vogel’s four years at the helm. By 2003, Charter’s debt load reached $20 billion. Much of it traded at less than half of face value, a dark forecast of the company’s solvency. As the stock price plummeted, the people who’d sold us their cable systems redeemed their shares. I was forced to pay out another $2 billion in cash.
That fall it came time to drink “some castor oil,” as I later told BusinessWeek. I fired Bill Savoy and pared my investments to forty companies, with an eye to diversification and maximizing return—basic principles that I’d almost lost sight of in my headlong pursuit of the Wired World. (One of those shed was Asymetrix, my first solo business. After a subsequent merger, it is now called Sum-Total Systems and continues to be a leader in e-learning software.) The irony is that I was forced into retreat just as events affirmed my vision. Broadband data penetration rose from 0.3 percent of American households in 1998 to 7 percent in 2000 to 61 percent in 2005. The digital platform I’d imagined was fast becoming a reality.
In August 2005, Charter finally got the CEO it needed in Neil Smit, a former Navy Seal and an outstanding executive from Time Warner. Over the next three years, we fought our way back to stability by extending maturities on our debt load, keeping our creditors at bay and our bondholders paid.
Neil ran Charter’s day-to-day operations with both expertise and a needed grind-away mentality. As he trimmed bureaucracy and got digital telephony running across our systems, fewer customers deserted us to satellite. Charter’s cash flow was growing 10 percent annually, and would soon catch up to our interest payments. It looked like the company might have a future, after all.
But our progress could last only as long as the credit markets gave us space to breathe. On September 15, 2008, Lehman Brothers fell into bankruptcy. The credit markets seized up, and commercial lending ground to a halt. Our refinancing options disappeared overnight. Charter’s debt load had swollen to $21 billion, with nearly $2 billion coming due in 2009, and the company lacked the cash flow to cover it.
In February 2009, after a decade of losses, we finally ran out of runway. Under the reorganization plan proposed by Lazard, Charter stockholders’ equity—including my 52 percent share—would be virtually wiped out, a bitter pill. But with the company casting off $8 billion of debt, junior bondholders would invest $1.6 billion in new capital. Charter would reemerge stronger than before.
The wild card was how the company would pay off $12 billion of senior secured debt held by JPMorgan and other big banks. If the obligations were reinstated at their original, low-percentage interest rates, the restructuring could move forward. But if they were reissued at steeper current rates, the bump in annual debt service—hundreds of millions of dollars—would sink the whole plan. Without liquidity, the junior bondholders would lose all incentive to convert their bonds into equity. The company might lurch into “free-fall” bankruptcy, with uncertain implications for all concerned, including sixteen thousand employees.
To avoid a free fall and get the loans favorably reinstated, Charter had to show that the same interests would remain in control before and after the reorganization. The junior bondholders needed my cooperation (and my retention of a 35 percent voting stake) for the agreement to work. I wanted to help ensure a smooth transition. But it also seemed reasonable that I get some consideration for my role in the plan, which would save the company billions in interest payments. We arrived at a compromise, and Charter filed under Chapter 11 in March.
JPMorgan led the court challenge. On November 17, the U.S. Bankruptcy Court for the Southern District of New York ruled in our favor and affirmed the reorganization plan. A month later, Charter came out of bankruptcy. I resigned as board chairman, and later Neil Smit moved on to become Comcast’s president of cable operations. We left behind a company with positive cash flow and a strong foundation, but the lessons I learned were among the most expensive ever. My net loss in the cable business was $8 billion.
WITH THE CLARITY of hindsight, I could say that I took the wrong people’s advice in plunging into Charter. I needed savvier, more experienced executives to assess my risks and to run the company, and I didn’t have them until it was too late. But the fact remains that the investment was mine, and I made serious miscalculations. Most of all, I failed to understand the downside of over-leveraging. My dreams of a Wired World empire finally sank under the weight of Charter’s mountain of debt.
In placing the biggest bets of my life on cable, I focused on its potential to change the world, not the downside scenarios. After embracing SkyPix early on, I underestimated the challenge from satellite systems, with their reputation for superior service. I did the same with the phone companies’ video and data offerings. And I failed to discern that the cable pipe couldn’t galvanize change by itself, or at least not as quickly as I’d thought. Operators have reaped handsome profits from selling high-speed data, but they’ve yet to capture much added value with new products and services over the top of the data stream. Even today, the fat pipe remains by and large a “dumb” pipe.
At the same time, recent trends suggest that I was more right than wrong in my prediction of a broadband future. As TV channels’ subscription fees squeeze profit margins, high-speed data is more vital than ever to cable’s growth. People have a voracious appetite for faster information flow, and fiber-optic technology has made cable far and away the top provider of digital data into their homes. The fat pipe has helped bring Amazon, Google, Facebook, and YouTube into near-universal acceptance. It has changed the way we live.
As the industry embraced DVR boxes and other services I’d been early in urging, operators found that customers were willing to pay for them. Now they can plow the extra revenue into the next round of innovation, in a virtual cycle.
The Consumer Electronics Show in Las Vegas each year contains a cavalcade of ideas that once percolated within Interval Research and my vision for the Wired World, from wearable HD camcorders and holographic displays to fully functioning multimedia set-top boxes and TVs, Web browsers included. In some cases, I was just too early. In others, our execution failed for a slew of reasons.
But even if I’d had more luck in my timing, the cable industry was wrong for me. In consumer electronics, product cycles run as short as six months; in computer software, about every two years. But in cable, it takes five years or more to introduce something new across the customer footprint. Cable is like a mule train. It’s moving as fast as it can but still takes forever to get anywhere. Case in point: When my company, Digeo, developed an Emmy-winning set-top box, it could not gain traction with Comcast and Time Wa
rner, which were wedded to inferior boxes from their subservient legacy suppliers. Only recently have they begun to mandate Digeo-caliber boxes for future deployment.
Or consider that cable has yet to fully incorporate mobile phone service, a platform that now supersedes landlines, in a triple or quadruple play.
Consumers who came of age in the digital era are agnostic about delivery systems. A bit is a bit, regardless of how it reaches them. Though many of us have broken our picks in the pursuit of interactive television (including the ill-fated WebTV), there’s no question that television and computer platforms are now converging. People might not choose to read long e-mails on their fifty-five-inch screens, but they’ll use a tablet as a TV remote while scanning their Facebook page. Or they’ll use Xbox to manage their photos and music.
An Internet port will soon be standard on higher-end television sets. If the cable industry doesn’t move aggressively to integrate Internet functionality with its TV offerings, providers like Apple TV and Google TV will fill the void with “over-the-top” services. If you can stream Amazon.com or Netflix films and videos to your television, with their tens of thousands of on-demand titles, how appealing is pay-per-view? More threatening still, broadband channels are moving to distribute “linear” network content as well: CBS, TNT, Comedy Central. Multichannel television could be gradually supplanted by streaming video, with only sports and political events left as obligatory real-time viewing.
The technology already exists. The consumers are up for grabs. The digital future will belong to those who seize it.
CHAPTER 20
SEARCHING
In the early 1990s, Carl Sagan met with me to pitch a cause he held dear. The federal government had been funding the SETI (Search for Extra-Terrestrial Intelligence) Institute through NASA, in what was supposed to be a ten-year plan to observe neighboring stars. But Congress balked, with one senator from Nevada calling the initiative “a great Martian chase,” and the appropriation was canceled. The search for a signal from outside our solar system was about to be shut down.
“SETI’s taking on one of the great scientific questions,” Sagan said. “We need someone to step in and save it.” He was delightfully sharp-witted and persuasive, and it didn’t hurt that I’d watched every episode of Cosmos, his classic PBS documentary series on the universe and man’s quest to understand it. Along with Gordon Moore, Bill Hewlett, and David Packard, I agreed to give $1 million to keep the SETI Institute running. It was just enough for the operation to pay for a bit of observation time on giant radio telescopes in Australia, West Virginia, and Puerto Rico. SETI was then looking at a mere 750 stars, a paltry number against the 200 billion in the Milky Way alone. To have even a ghost of a chance to succeed, it seemed clear that a dedicated telescope was needed.
Several years later, researchers figured out how to process data from “ganged” small-dish radio antennas, a big breakthrough for radio astronomy. The idea of creating the world’s best SETI telescope—at a fraction of the cost of a single large dish—was enticing. I underwrote the creation of an installation at Hat Creek Observatory in the Lassen National Forest in northeastern California. Nathan Myhrvold, the former chief technology officer at Microsoft, chipped in for an electronics laboratory at the site. In 2007, after years of research and development, the Allen Telescope Array opened its “ears”: a set of forty-two six-meter dishes combing the sky in a thorough, methodical hunt for a signal that might change everything.
The telescope array works on the principle that objects in space emit radio wave “signatures” that describe their size, shape, and chemistry. Much longer than optical waves, radio waves are less scattered by space dust and can get to us intact from the edges of our galaxy and beyond. When SETI uses the array to search for a signal, it can scan ten times more of the radio spectrum than any previous installation. Its detection beams can focus on six stars at once, or on three stars with two simultaneous beams apiece.
The array’s other big advantage is its wide-angle, high-resolution view, which captures a field of sky as large as seven full moons across. (The thousand-foot-diameter Arecibo Telescope in Puerto Rico is far more sensitive, but it looks at a small area of space through the equivalent of a soda straw.) SETI has compiled a list of a quarter-million sunlike stars, the ones most likely to have livable planets within six hundred light-years of earth. In five years or so, after the galactic census satellite Gaia begins sending back its survey data, that catalog will swell to several million stars, a decent foundation for this type of search.
What’s more, the array is a Moore’s-law telescope; its digital signal processing will keep improving exponentially. It’s already 100 trillion times more capable than the one that SETI founder Frank Drake used when he started signal hunting in 1960. The Institute’s goal is to expand to 350 dishes, which would make the array one of the more powerful radio telescopes in the world.
Though there are no guarantees that SETI will turn up an alien communication, the history of astronomy suggests that its new-generation technology may lead to unexpected discoveries. With a portion of the array’s observing time, the University of California–Berkeley’s Radio Astronomy Lab is conducting investigations in more conventional astronomy: gamma ray bursts, black holes, stellar explosions. By mapping our galaxy’s distribution of hydrogen gas, an essential ingredient in star formation, the Berkeley data should give us a clearer picture of the nature of dark matter, the galactic life cycle, and the structure of the cosmos itself.
WHENEVER THE ARRAY finds a SETI “candidate signal” that stands out from the background of garden-variety electromagnetic noise, a gauntlet of tests winnows out false positives. Computers quickly determine whether the candidate came from the scanned star or an orbiting satellite—or a stray cell-phone crackle. Once a signal successfully passes those tests, other radio telescopes will be contacted for independent confirmation. And if and when SETI actually verifies an engineered communication, a tweet across the cosmos, I’m told that I’ll be the first person that director Jill Tarter calls outside her professional community.
My phone hasn’t rung yet, and there’s no way of knowing if it ever will. Frank Drake devised an equation that can theoretically calculate the number of communicating civilizations in the Milky Way. But because we can’t determine the Drake equation’s parameters (such as the life spans of civilizations that develop broadcast-capable technologies), it’s hard to know what the real probabilities are. If those civilizations last only a few thousand years, the SETI Institute may be out of luck. If they last a few million years, our odds are far better.
When it comes to the existence of extraterrestrial life, there are strong arguments on both sides. In Rare Earth: Why Complex Life Is Uncommon in the Universe, Peter Ward and Donald Brownlee suggest that the specific conditions that produced animal life on earth—our distance from the sun, the amount of water in our atmosphere—add up to an unlikely accident. Yet recent research has shown that cellular organisms can exist at more extreme temperatures than we ever thought. Nearly five hundred exoplanets, those belonging to other stars, have already been discovered, and the Gaia probe should find tens of thousands more. In theory, any one of them could be the winning lottery ticket.
SETI is the longest of long shots, but I find its question gripping. Do we have company in the universe, even in our own galaxy? A yes would have implications we can only guess at. Any society with the ability to signal its existence would almost surely be older and wiser than we are, with technology that might offer huge benefits. But even if we never made contact (or followed Stephen Hawking’s recent warning and declined to return the call), a confirmed signal by itself would permanently alter our perception of the universe.
IF SETI REPRESENTS our outward search for intelligence, a Vulcan initiative called Project Halo is helping to lead the inward search: to design software that can simulate certain aspects of human thinking. What we now refer to as artificial intelligence, or “AI,” dates back at least
to 1921, when a Czech science fiction play called R.U.R. coined the term “robot.” When I was young, HAL-9000 (in Kubrick’s 2001) and Colossus (from the novel and movie of the same name) embodied nightmare scenarios in which super-intelligent computers turned on their human masters. Machines that behaved like people, even people gone mad, were all the rage back then.
But for me, even more compelling was the sci-fi theme of a dying or threatened civilization that saves itself by finding a trove of knowledge. Tagging along with my father to his library job, I spent hours amid acres of shelves that held what seemed like an infinite mass of information. The idea of gathering all the world’s knowledge in one accessible repository—like the Final Encyclopedia in Gordon R. Dickson’s classic of that title—seemed both grandiose and seductive, with untold benefits for humankind.
With the development of the World Wide Web in the nineties, there were glimmers of hope that this repository might be under construction online. In reality, though, the “knowledge explosion” left us with mounds of sources but no direct way to get a quick and concise answer. It became too easy to get lost in a tangle of text and hyperlinks. And while modern search engines have proven to be invaluable in presenting lists of pages with specified keywords, they still fall far short of the ultimate goal of software that understands.
Aristotle, the Greek scientist and philosopher, was literally a know-it-all. He mastered the knowledge of his day on every topic that mattered, from history and political science to medicine and physics. Even more impressively, he could explain what he knew to his students. But in today’s world, where scientific knowledge may be doubling by the year, it’s impossible for any one person to absorb more than a small fraction of it.