[The following is the speech text for the keynote I gave at the SwitchPoint Conference April 20, 2012]
It is sometimes said that computer scientists worry about only three numbers: 0, 1, and N, where N tends to get very large. Sometimes such oversimplifications can lead to astonishing insights, such as the one that I had 25 years ago in June of 1987.
Do you remember 1987? Greed was good, junk bonds were king, and zero was the biggest and most important number on Wall Street. Zero drove all the arbitrage equations, because both sides of the arbitrage are supposed to sum to zero. Arbitrage is a special case of the zero-sum game, a prominent theory promulgated by all respectable business schools of the day. Zero-sum logic made it a moral imperative to ensure that success was not just about winning, but about making sure that everybody else lost. You were not forced to like the terms of the game, but you were damn sure forced to accept them as they only way to play.
I didn't accept those terms for two reasons. First, I had been rejected by every respectable business school I applied to. Second, I realized that zero was not the only number in the playbook. N, especially when N becomes very large, is an even more interesting number.
Consider that every year the Earth receives 120 Petawatts of energy from the sun. That's 120 followed by fifteen zeros; it's also 8,000 times what we humans generate at our current industrial scale--120 x 1015 a really big N! It's so big, in fact, that thinking about life on Earth as a zero-sum game just makes no sense. This positive energy flux not only powers life itself, but it represents an enormous flow of available possibilities. As soon as I realized that we were living in a positive-sum world, at least when it comes to energy flux, I realized that the strategies for "winning" the zero-sum game and the positive-sum game were quite different. We win by having more winners, not by increasing the number of losers. What a different game that is! And what an opportunity to win!
It is here that I should give some credit to those who taught me mathematics, logic, and theory. Perhaps you have read Gödel Escher Bach by Douglas Hofstadter, which includes the very accessible story of how Kurt Gödel managed to prove that any system of logic powerful enough to prove a statement true or false is necessarily incomplete (meaning there are true things that cannot be proved) or inconsistent (meaning that there are true things that can nevertheless be proved false). If you have, you know that Gödel takes some amazing leaps in directions that make almost no sense until he brings them all together in a logical tour-de-force. Perhaps it was that very story that gave me the confidence to not worry at all about the individual propositions of the positive-sum game (which, by the way, are each absurd in the zero-sum context), but rather to let them all take shape as a community of ideas, and to be prepared for an astonishing result.
The answer I derived happened to agree perfectly with The GNU Manefesto, a polemic written by Richard Stallman, founder of the Free Software Foundation and creator of the GNU Project. Stallman railed against the injustice of software hoarding; I calculated that excluding valuable participants from an effort was self-defeating. Stallman talked about the moral imperative of sharing; I saw free redistribution as a frictionless way to build market share. Stallman talked about the vital importance of community; I realized that a diverse marketplace of people having problems and people solving problems would lead to more rapid and more relevant innovation. As I recently posted on Google+, I was so sure that this way of thinking could lead to a $1B company that I started Cygnus Support in 1989 just to prove it. Red Hat acquired Cygnus in 2000, and twelve years later grew through the $1B mark in convincing fashion. But being first to $1B does not mean we are the only winners. The founders of Facebook, Google, and Twitter have all said that they could never have started their companies without open source software. They are also winners who recognize the power of open innovation. And open innovation reaches far beyond software.
Harold Varmus shared the Nobel Prize in medicine in 1989 for "discovery of the cellular origin of retroviral oncogenes." Varmus went on to lead the NIH, dramatically increasing its funding from $11B per year to $16B per year during his appointment under President Clinton. I met Dr. Varmus in Kyoto at the Science, Technology and Society summit in 2006. He told me that of all the things he did after winning the Nobel Prize, the one he thought was most important, because of its long-term and wide-ranging impact, was to spearhead open Internet access for medical research. In 2007 Varmus summarized his inspiration in a speech, quoting Panizzi (the librarian of the British Museum from 1837-1866):
"I want a poor student to have the same means of indulging his learned curiosity, of following his rational pursuits, of consulting the same authorities, of fathoming the most intricate enquiry as the richest man in the kingdom, as far as books go, and I contend that the Government is bound to give him the most liberal and unlimited assistance in this respect."
As Varmus explained to me, the instant he saw the World Wide Web, he recognized that open access publications could be deposited in an online public repository immediately upon publication and also offer access to the widest array of users possible. Varmus saw what I, too, saw in my own experiences with open innovation: the advancement of information and communication technology coupled with the economies they provide led to a fundamental shift in both human productivity and, more importantly, the capacity of the community. It did not matter whether the information came from the analog world or was born digital: once the information was infinitely available at virtually zero cost, the rules of a whole new game were in play. For Varmus, this meant that digital libraries could "rise above the strangulating effects of [the] so-called 'Gutenberg liabilities' inherent in traditional scholarly publishing."
In 2010, the Public Library of Science, founded by Varmus, Patrick Brown, and Michael Eisen in 2000, received more than 20,000 scholarly submissions, publishing nearly 10,000 with an operating budget of approximately $13M, which is very close to the list price they offer of $1,500 to have a paper properly reviewed and published. That may sound like a lot of money compared to what newspapers charge for classified ads, but contrast this with Elsevier, which publishes 250,000 articles in 2,000 journals with revenues of $3.2B and a profit margin of more than 36%. Elsevier averages nearly 10x the cost, but remember: cost-per-article alone is not the only problem that Varmus was trying to address. An article from WIRED magazine explains:
[Patrick] Brown and a fellow of his, Michael Eisen, were using a technique for genetic analysis that benefits particularly from information-sharing across many labs. Eisen had written a program to enable such exchanges, but it was constantly banging up against copyright restrictions from journals that didn't want the data in their articles passed around. Brown and Eisen thought that was insane. For a decade, the physics community had run an electronic archive out of Los Alamos National Laboratory, where researchers could upload their papers even before having them formally accepted by journals. The physicists loved it. Why not create something similar for the bio-medical community?
The conversation clicked with Varmus. Soon, he, Brown, Eisen, and David Lipman, an NIH colleague, drafted a proposal for "E-Biomed," a public, one-stop repository of biomedical papers. Researchers could submit papers through a peer-review process similar to that used in the current journal system, or they could post work fresh out of the test tube.
E-Biomed evolved, and today in addition to the Public Library of Science, PubMed, PubMed Central, 38 other databases are all cross-linked, many in a multitude of dimensions, now provides virtually all biotech information known to the NIH through a single, easy-to-use open source interface called Entrez (ftp sources here). Several of these databases have more than 10M entries and many more have more than 1M entries. Just imagine all the connections between all these datasets, and how that informs what we know and what we wish we knew about how our own source code actually works. And imagine how powerful it is to have the freedom to write programs that can traverse all this data and develop new knowledge and connections. I have spoken with numerous heads of other nation's medical research facilities, and for all of them, the NIH's open access policies are the gold standard around the world. The openness is the innovation, and the numbers show it: at last count more than 1M researchers a day ran an annual aggregate of 1.8B searches against MEDLINE/PubMed and downloaded 4TB of data per day from the National Library of Medicine's servers.
But not everybody wants to play by these rules, and in true zero-sum fashion, they want to prevent us from playing by these rules, too. A decade ago, Varmus tried to make peace with the other side by offering to delay public access to results published in traditional journals for a period of six months--a bitter but pragmatic compromise in his view. Bitter because Benjamin Franklin could regularly correspond with his fellow scientists across the Atlantic every 6-8 weeks. You would think that with the Internet and 250 years of technology advances that scientific dissemination would speed up, not slow down. But alas that was not enough for those who place profits ahead of progress and monopoly—the very enemy of democracy—above community welfare. In January of this year, lobbyists for the traditional publishers helped draft the Research Works Act, which, among other things, stipulated:
No Federal agency may adopt, implement, maintain, continue, or otherwise engage in any policy, program, or other activity that:
(1) causes, permits, or authorizes network dissemination of any private-sector research work without the prior consent of the publisher of such work; or
(2) requires that any actual or prospective author, or the employer of such an actual or prospective author, assent to network dissemination of a private-sector research work.
Needless to say, such a bill would not only end the NIH's Public Access Policy, but it would forbid any effort on the part of any agency to ensure taxpayer access to work funded by the federal government. That bill has been killed once, largely by the same wave that sunk SOPA and PIPA, but you know as well as I do that bills in Congress are like zombies and we must remain vigilant. We're the ones spending $28B per year on medical research--we have a right to share in the data we paid to develop!
Let me close by taking a step back and competing the two games I have presented: the zero-sum and the positive-sum. Which game itself is the winner? The finite game, where every resource, every move, must be subtracted from one column before being added to another, or the infinite game, where multiple possibilities co-exist, either cooperatively or independently? The closed game, where you can be excluded from playing, either because you don't know the trade secrets necessary to play or because you don't have the necessary rights to practice patents or share in the fruits of what has already been paid for? Or is the open game better, where all the rules are published, and all who want to learn them can play?
This is where the number 1 comes in. Gandhi famously said "Whatever you do will be insignificant, but it is very important that you do it." Each of you is the 1. I am the 1. When one person can successfully add energy to the efforts of the community at large, that person augments and amplifies the power of the larger community. In the positive-sum world, everything we do adds up. I am just one person, but I've put 25 years into open source software, and I've seen other creative, open, collaborative communities emerge: Creative Commons, the Science Commons, the Conservation Commons, and more. One by one, day after day, year after year, we have created billions of lines of software source code, hundreds of millions of creative, copyrightable but also shareable artifacts, and we are just getting started. I can't wait to see what the next 25 years can bring!