The Panton Fellowships applications will close in 9 days, on 24 February. You should probably apply if you’re reading this blog and you’re a grad student or a budding scientist. Adapted from the OKFN blog: The fellowships are for scientists who actively promote open data in science - awardees receive £8,000 over one year, plus a small discretionary budget for travel and related expenses. They are designed to be flexible, and there is scope for Fellows to carry out a wide variety of activities. Applicants are encouraged to propose their own work plan. Panton Fellows may wish to initiate discussion about the role and value of openness, explore practical solutions for making data open, and push for change in scientific practices. Panton Fellowships are open to all applicants, and are particularly suited to graduate students and early-stage career stage scientists. I was an original author of the Panton Principles around which the Fellowships are centered, so I have a bias towards getting awesome people into the application pool. But the reality is that a Fellowship like this has rewards that go well beyond the monetary ones, which are honestly and obviously limited. Panton Fellows will have wide recognition in the community, and a chance to work with some of the leading minds in open science and open data. Give it a shot.

Seriously, I wish I could get cranked about yet another piece of “please protect my content industry from teh internetz” lobbying crap, but the only reaction I can really summon at this point is LOLCats and animated gifs of Bill Cosby.

I spend a lot of time ranting on about openness, and a lot of that time I am preaching to a choir or people who believe in openness at a deep enough level that they don’t think about the why of it. It’s just accepted that open = good in some epistemic sense. And that’s nice in a lot of ways. It’s nice to be able to just talk about good stuff and not make basic arguments. But I arrived at openness from a different place than a lot of my compatriots. I didn’t arrive at it from a political freedom perspective like the founders of the free software movement did, although I now embrace the political. I didn’t arrive at it from a horror at copyright expansionism like the founders of Creative Commons did, although I now share that horror. Nor did I arrive at openness after experiencing the misery of scholarly publishing models like the founders of the Public Library of Science did, although I now know that same misery. I arrived at it after getting my nose bashed in by biology. I started a company about thirteen years ago, based on some work I did in university epistemology and semantics, whose entire premise was that we knew enough about biology that if we simply converted that knowledge to formal ontologies and then used common identifiers to connect those ontologies into networks, we’d discover some basic laws of biology that were implicit in the knowledge but invisible to us because of format problems. Was I ever wrong. But I was wrong in a way that is a leitmotiffor people entering biology from other fields, whether technologists or philosophers, whether naive twentysomethings like me or successful Silicon Valley executives: I bought into reductionism.

So this happened yesterday, and it’s a good thing. But note Mike Eisen’s question below Ed Summers’ tweet.

I have been obsessing lately with how many people we’ll need to recruit into the PLC-CGR study to make a real difference. On a good day I think it’s 100,000. On other days I think it’s well into the millions. But it’s gotten me thinking about magic numbers. Every “open” community has a magic number - the number of active contributors that make a project into something that lives and breathes. As Nathan Yergler has pointed out beautifully, “open source” is not a verb. It’s about communities of people who do (often boring) work, willingly, for a multitude of reasons, to create open stuff. And there’s not a lot of evidence that people actually want to control their own health information in a way that gets PLC-CGR to its magic number. Indeed, some of the evidence points the other way. Wikipedia’s magic number seems to be 300,000 - that’s how many editors make edits every month. Of those editors, 50,000 make more than five edits per month, and 5,000 make more than one hundred edits per month. That puts the magic ratios at 60:1 for serious editors and 6:1 for occasional editors. And that is all out of about 16.5 million Wikipedia users who have bothered to sign up for an account, which is a lot less than the actual total users who’ve ever looked at Wikipedia. One of the reasons I’m so skeptical of “open source science” as an approach is the magic number problem. The total number of people who know enough stuff about any given topic to edit Wikipedia is so vast that even a tiny fraction gets you to the magic number. But when we’re talking about medicinal chemistry, those same percentages probably don’t even get you a single postdoc. We often (I did, for a long time!) assume implicitly that scientists are somehow more open than society at large to the ideas of knowledge sharing, and thus the math will be better. I’m not sure. But the PLC-CGR study routes around this by recruiting normal people, not scientists or doctors. And it doesn’t require wikipedia level editing commitment. But in an ideal world, we’ll have a hard core of people who don’t just create an account and never come back. Who don’t just upload their genotype from 23andme once and never come back. Who don’t just upload a crappy EMR in Blue Button and never come back. We need a group that decides their free time is well spent donating their data to science, who are willing to RTFM enough to get not just today’s data, but tomorrow’s data, into the system. To serve as community moderators and admins to help the newbies along. That’s hard enough. But on top of that, we need that group to be large enough that the science that emerges is replicable, statistically sound, and not horribly overfitted to the sample we provide. I’m more optimistic than the Health Data Management article. I think one of the reasons for the small numbers of genotypes bought, or the small numbers of people controlling their own health data, is that there isn’t a route for that data to actually turn into results. I downloaded my 23andme file and it’s useless to me. But if we know that we can funnel our data into a system that gets it to researchers, then we increase the motivation - at least by a little - to take some control ourselves. If we knew the magic number, this would all be easier. But then again, if we knew the magic number, it’d probably already be done.

I’m speaking today on a panel at the Atlantic Live healthcare forum. Below are the themes I’m going to hit in my talk - but it felt like it was worth posting them here as well. We have engines to generate correlations out of massive data that are unreasonably effective. We’re not using them in health, because the machines don’t work very well on health data. It isn’t just that the data be massive, but that it be available (i.e. open) and at least somewhat standardized. Health data’s almost never either one. Here’s a key sentence from the Google paper linked above: “With a corpus of thousands of photos, the results were poor. But once they accumulated millions of photos, the same algorithm performed quite well.” Have we ever had millions of data records about health for algorithm training? Nope. That’s a tragedy. It’s not that these engines - machine learning - are the magic answer. But they’re an incredibly powerful toolkit for finding correlations, which helps us decide what experiments to run to test for causation. It’s like we’re carpenters and we’re sitting around building a house and saying, nah, we don’t want to use power tools. Hand saws for everyone! The Amish approach won’t scale. We have no idea how many people it’s going to take, or how much data about each person it’s going to take, to discover exactly how valuable machine learning is going to be in the health space. But that’s a terrible excuse to do nothing. In fact, it should be the spur that gets us started on the experiment itself. That’s what i am doing with the Portable Legal Consent project. It’s an experiment to test three things. 1. can we use a standardized approach to informed consent that disintermediates the traditional study systems, which are crusty and have insanely high transaction costs? 2. will people enroll and upload data, and of what type and quality? 3. what kind of results will emerge from the computational research? Until we know these three things we’re basically arguing about philosophy and not reality. We know that these systems are tremendously powerful and predictive in many areas. We know that we aren’t using them in health in a meaningful way. And we know that there’s nothing stopping us from trying except ourselves.

Just a quick hit to link to Daniel Solove’s epic paper “A Taxonomy of Privacy” which has rapidly become a major influence on my work at Consent to Research. It’s a free download. But the first sentences of the abstract are a great teaser: Privacy is a concept in disarray. Nobody can articulate what it means. As one commentator has observed, privacy suffers from an embarrassment of meanings. Privacy is far too vague a concept to guide adjudication and lawmaking, as abstract incantations of the importance of privacy do not fare well when pitted against more concretely-stated countervailing interests. Agreed. I love the taxonomy’s distinction between information collection, processing, and dissemination. I do wish there was a discussion of harm in here because economic harm is a real consequence of information dissemination in health.

I come back from vacation and nearly have an aneurysm. Seriously, someone thinks we should extend the HIPAA model to consumer privacy. Because it works so well. Encourages so much innovation. Protects consumers and customers so thoroughly. Let’s use one of the worst-working data systems we have as a model! It is depressing how often we in the “open” world have to say this, again and again. More control via rights is not the answer to every problem. Not in copyright. Not in patents. And not in privacy for data.

In my spare time when not working on Portable Legal Consent, I continue to work on Open Access issues. About ten days ago, I got home from vacation to news that I could go to Washington DC and meet with John Holdren, the Science Advisor to President Obama. He’s a nice guy, and it was a great meeting. He seemed to understand the issues and our points. Like any good political appointee he was non committal of course. But it was a brutal trip. Redeye on a Tuesday night from the West Coast after a full day of meetings, change into a suit (and shave) in a public bathroom because Dulles Airport has no showers, meeting, back onto the plane home. And it hit me - us, because I was with Mike Carroll, Mike Rossner, and Heather Joseph - that the redeyes and the meetings and the arguing were not carrying the day. We needed to do something else. So we started the Access2Research campaign to engage the public in open access. Please go, read the context, and if you agree, sign the petition. The only thing missing from the open access debate is the public. You can remedy that - but you’ve only got 30 days to sign, and we need 25,000 signatures to carry the day.

The Access2Research We The People petition has been a far greater success in its first week than I at least dared hope, when we hatched this plan and took this picture:

I’ve been working up to a rant on twitter lately about Open Access, as the Access2Research petition’s first week drew to a close (sign it). The usual suspects are making the usual arguments - OA will dictate where scientists have to publish, OA will kill peer review, and most offensively to me, science is too complex for us unwashed liberal arts heathens to possibly understand, so no good will come of access. But science is a place where we keep out the unwashed masses. We no longer credential computer scientists (well, universities churn them out, but your credentials on github matter a lot more to savvy programmers than a CS degree from a state university - you’d be better served majoring in something fun and checking in code to open source projects). Has innovation in computer science been a problem? [crickets chirping] Basically every knowledge based discipline that runs on digital content has been transformed. Software. Journalism. Music. Video. And you can track the innovation patterns of each one based on the level of control that institutions maintain. Note: not all innovation is useful - most of it is shit - so part of my argument is that radically increasing the rate of *all* innovation is the best mathematically certain way to increase the rate of *useful* innovation. It’s like art. Most art sucks. But if enough people make art, then even if the rate of awesome artists doesn’t improve, making more people overall be artists means more awesome art . That’s what’s happened in software. More people make it. That means more shit software. We just don’t use it (ever browse the Android app store’s dregs? Sheesh). It’s happening in journalism, whose business model turned out to be based on classified advertising and got eaten by Craigslist, the ugliest website on earth. It’s happening in music, where Apple ate the music industry’s lunch, where artists can raise a million dollars on Kickstarter just as their old labels go bankrupt. But science isn’t like that. Science is a lot more like the cable industry. Comcast and a few behemoths control the last mile of the internet to most houses, and so we don’t even realize the world we live in is radically limited. Internet in the US is so bad compared to so much of the world and we don’t even see it. Toll access publishers of science are just like Comcast. They want to control the last mile. And scientists who buy into the argument that those of us in our houses, lacking credentials to understand their science, are perpetuating a knowledge lockup. They’re on the wrong side of history. You see, it does not matter if 999 of the 1000 people who read an open access article, who might not otherwise have been allowed to read it without paying $50, fail to understand it, believe they have disproved the second law of thermodynamics, etc. It matters that the one person does read and understand is provided access. Because then, in that moment, we’ve created a scientist - or at least the makings of one. And the only people that threatens are those counting on their credentials to keep them competitive, or profitable, or employed. Since I’m none of those three it’s pretty easy to support open access.

An old friend of mine asked me, after seeing my torrent of posts about the Access2Research petition (go sign it PLEASE) asked me this: What the heck is this all about? I hate the fact that the Internet people can use cookies to spy on me, track my computer use and invade my privacy. Will your open Internet protect their ability to continue to violate my privacy? This is a really old friend. We met 22 years ago, in school. We took introductory philosophy together and we were jackals in that class. We lived across the hall from each other the next year. He’s one of the smartest people I met in school, and a genuinely good guy. He actually wrote letters to his friends, including to me, before there was Facebook. He introduced me to online networks - I first got on the internet via his Mac, onto AOL, in 1991. The fact that we have to explain open internet to my friend is part of the problem we face. Indeed, it may be the most important part of the problem. So, old friend, here’s why I think this is important. The internet itself, and the web itself, are “open” at their core. That is to say they are built on open technical standards that define how computers send bits across telecommunications networks (the internet) and how computers send and reassemble and display documents across networks (the web). No one can “own” these standards. No one can lock out anyone who wants to add a computer, or add a web page. And anyone can *add* stuff to both the internet and the web - including security, like https (on which Amazon transactions thrive securely) or social networks like Facebook. You need never ask permission to add things to open systems. The market decides if they work or not. The openness of the systems is at their core. The thing is, building closed systems on top of the open internet is awesome business. Thus, companies build browsers that track all your behavior (sometimes they even offer you small cash prizes!). Facebook tracks you while you’re logged out. The only reasons browsers are even contemplative of “do not track” is because the competition from the open - Firefox - is so strong. In the absence of competition from the open there is just exploitation. That’s what Facebook’s all about. I use Facebook. The question was posted on Facebook. It’s a great way to communicate with my family and old friends. But the lack of competition from the open allows Facebook to make choices in a lopsided market, ones that we don’t have to worry about when it comes to the internet. Because Facebook can delete your page at will. Facebook can shut down your app company at will (this applies to Apple as well). There’s a czar in a closed system, whereas in an open system, the center is shared openness, what William Gibson called a consensual hallucination. So in open systems, it’s actually both more likely that someone will build tools to exploit your privacy (because it’s good business and has already happened) and that you’ll be able to resist those tools (because there is competition from the open and you’ll be allowed to install resistance software). In a closed world, only the former takes place. I see open access to the scholarly literature as part of that continuum. Freeing knowledge from behind artificial walls makes knowledge an open system like the internet or the web. That’s going to mean a lot of the scientific and scholarly equivalent of cat pages and blink tags and all the bad stuff that went along with the early web. That’s going to mean a lot of pseudonyms and trolling like we have with today’s web. But it also means that good businesses can be built on top of the knowledge web, and that as those businesses inevitably look to monetize, we have the freedom to resist the less savory choices they might make. Open systems are about freedom. Not just of cost, but in the constitutional sense. Sign the Access2Research petition. Please.

The Scorpion and the Frog is a fable about a scorpion asking a frog to carry him across a river. The frog is afraid of being stung during the trip, but the scorpion argues that if it stung the frog, the frog would sink and the scorpion would drown. The frog agrees and begins carrying the scorpion, but midway across the river the scorpion does indeed sting the frog, dooming them both. When asked why, the scorpion points out that this is its nature. The fable is used to illustrate the position that the behaviour of some creatures is irrepressible, no matter how they are treated and no matter what the consequences. (from Wikipedia) Lots of folks are getting hot and bothered over the 23andme announcement that it has been granted its first gene patent. It’s caused a lot of chatter on Twitter, and a lot of handwringing. Let me ask you all something. Why on earth did you expect anything else? Genotyping is a commodity service. That’s not 23andme’s business. Their business is selling the anonymized data to those who wish to use it for research purposes  and in doing their own research on the data. They tell you this up front, as they note in the comments on the blog post: We make reference to our intent to pursue intellectual property rights for discoveries made from our research in both Terms of Service (in section 13) and in our Consent document (sections 3 and 5). Companies exist not just to provide you with neat services, but to make money. And patenting genes is part of how companies in the drug and health space make money. Whether or not it leads to drugs is another question, one we might ask of shareholders of Millennium and Human Genome Sciences, two leaders in gene patenting yet not two leaders in actually curing very much at all in humans. Open systems matter because the core of the system is a shared set of norms. Sage Bionetworks is part of an open personalized medicine revolution. It’s why I’ve downloaded my own copy of my own data, and am uploading it into Sage’s Synapse compute environment, so that it can be part of fully open research. 23andme wants to be at the center - the czar - of a closed personalized medicine revolution. They want to build a walled garden so big that no one person notices the walls - unless that person wants to do something without getting permission from 23andme, like new research, or starting a company. That doesn’t make them bad, or their service bad. They’re the scorpion, we’re the frog. Hopefully the sting isn’t bad enough that we all drown. I want 23andme to make money. I’m a paying customer (indeed, I pay for my whole family’s accounts). I’ve filled out every survey, answered every snippet. But I don’t trust them to operate openly, or with my interests at heart. They have $50M in investment to recoup, indeed, to add a zero to. They’re going to make choices as they monetize that I don’t like. This is one of them - it’s both not likely to lead to a drug, and it pisses off their fan base. But I’m not remotely mad about it. It’s in their nature.

Giving a talk at Designing Geopolitics 2. Based in large part on ideas in my post on designing for emergence. Since I’m speaking in front of the optiportal, I have no powerpoint, but instead a folder of files that I’m using as a mosaic. I can’t redistribute all of them, so I’m not able to show you everything. Will add a link to the HD video. These are not my comments as they will be delivered. Think of them as chord changes. We’ll see if it works. 0 (displaying a giant photo of Copernicus) knowledge from data Copernicus. Photo. cultural object. Creative work. But since it’s digital, it’s also data. It can now be effortlessly copied, distributed, at zero loss of quality or resolution. And Copernican theory led to Brahe, who made painstaking observations, which allowed Kepler and galileo to figure out planetary motion. 1 (displaying a mosaic of pictures of old data) they did this using what they called observations, what we call data. this is what data used to be. it reformed geopolitics (heliocentricism) and transformed knowledge distribution. 2 (displaying a mosaic of paleofuture slides) Remember that with the data we have at the moment, we ususaly get the future hilariously wrong. Future shock images. Futurism tells us more about now than the future a lot of the time. Thus any geopolitics needs to understand that we are at least mostly wrong. 3 (display images of analog data storage and copying tech) these are all ways that we wound up storing data, from the old to the last of the pre digital. quality degraded with copying. all analog. all kind of human: liable to age in ways we understood, intuitively (get images) 4 (display a large picture of the @ sign key on an old typewriter) Transition to digital and what it means. 5 (display pictures of data capture technologies and shots of pigpen from peanuts) capacity to capture data exploding. faster, better, smaller, ubiquitous. pigpen secreting dust as we secrete data. 6 (display pictures of the Dust Bowl era) dustbowl of data. dust bowl emerged from a national politics of exploitation of the land, coupled with an unusually rainy period on the great plains. “Rain follows the plow” was the idea. Devastating impact on humans, on economy, on land. 7 (black screen) not to mention that digital data, though of the human, is distinctly digitally not-necessarily-human stuff. 8 (wipe through a set of pictures of a greenhouse monoculture, the earth at night from space, black box.) monoculture emerging. that is the geopolitics we are getting. digital divide emerging. that is the geopolitics we are getting. the black box transaction. that is the geopolitics we are getting. 9 (black screen) but, the trick. we can build our own geopolitics without asking anyone else’s permission. 10 (display a set of soviet propaganda poster) it is commonly built, but it is not sovietism. 11 (display photo of pic saying no public access) it is a reaction to the increasing privatization and exploitation of digital assets. no public access is the default rule. access might be free of cost, but you trade your inner pigpen for access. 12 (display photo of public footpath) common public access. free software. creative commons. science commons. where commons based systems work, where they don’t. property rights and their limitations, privacy rights and their implications. 13 (display photo of one green shoot in a furrow) the consented commons as the green shoot growing. it does not take all of us, even very many of us, working together to create something amazing. that’s the real lesson. there is already an asymmetry between the number of people who make things and the number of people who consume them. when we invert the system through a shared open core we can benefit from that same asymmetry to make open systems deeply competitive with closed ones. that’s why there is resistance to the open. 14 (display pics of wright brothers at kitty hawk.) where we are now: first flight at kitty hawk, first experiments. the older i get the earlier it seems. we’re learning, trying, hacking, failing. that’s fine. that is indeed the point. 15 (display pics from NASA explorations) the blurring of the lines: jupiter photo (get more) - are they art? are they data? does it matter? 16 (display the following words across entire screen, white text on black background) geopolitics of data as the potential for emergence.

I’m a relatively new dad. This is my second fathers’ day, but my son was so new last year that June is lost somewhere in the fog that comes along with four hours of sleep a night for weeks on end, 18 diapers a day, and feedings every two hours. It’s the first one I can really enjoy. Being a dad is a funny thing. I didn’t expect it to be this much fun . The diapers and the feeding and the repetitious activities are the things that so many people focused on before I had a kid, but very few people talked to me about how having a kid can drag you into the moment as well as, or better than, anything else. Watching the boy play on the patio in the warm afternoon sun. Giving him a bath. Helping him climb. Smelling his head in the morning while he takes his first bottle. I spend a lot of time - a LOT - plugged in and online and digitally present. My son tethers me to the real world. And he makes me appreciate my own dad more and more. I find myself unconsciously doing things my dad did. I make a silly face and catch myself in a mirror and it’s my dad’s face. I see his face in my son. I hear his voice in my voice when I talk, feel his structure in my structure. Happy fathers day, pops. Thanks for teaching me.

I’m at TED Global this week. I speak on Friday morning, closing the first session of the morning, just after Imogen Heap. It seems trendy in some of the circles that pop through my Twitter feeds to be cynical about TED. I never really bought into the hype in the first place, so I didn’t have anything to be disappointed about. But now that I’m here, and through a few days of it…there’s something to the hype. It’s a truly unique and immersive experience. You dance at 9:30AM to an Indian rock band. You see a guy who turns sound into paintings, because he’s colorblind and has a device that converts color to sounds for him and vice versa. You see *zero* laptops out during sessions. You get exposed to a diversity of ideas that simply doesn’t happen at other conferences. Everyone is open, everyone turns around and talks to you, no one turns away from an introduction. It’s different . I don’t have the money to go to TED unless I’m speaking. But I’m grateful to have the chance to be here, now, to be part of making this conference, to understand why it’s become a big deal. Yes, there’s precious stuff here. And I won’t lie - I’ve heard the word paradigm too many times. But I’m not going to pretend that being able to walk up and get an iced tea brewed from coffee cherries and dosed with elderflower syrup isn’t awesome. Haters, feel free to hate. I’m going to keep drinking the cascara.

My TED Global talk from yesterday has generated attention, as I’d hoped. I want to clear up one point from an otherwise excellent summary of the talk from the TED blog. The title of the talk could well have been the “unintended consequences” of informed consent. I’m not criticizing the idea that we should be informed before we join medical studies - indeed, in the talk, I note that this is a beautiful invention of society, something we should be proud of. My problem is that we’ve joined the idea of being informed with an idea of privacy  that is creating a drag on innovation. That prevents reuse of data without extensive permissions. That makes it hard to donate your data to science. Consent to Research is about untangling the two. It’s built on informed consent, not an attack on it. Because the reality is that privacy and consent have become entwined, and are used by researchers who are competitive, or just plain lazy or arrogant about our ability to understand data about our own bodies, to keep their data secret. Just as free software is a hack built on copyright, open consent is a hack built on informed consent. Simple as that.

Re-reading Holy Fire by  Bruce Sterling. It’s probably been a decade since I first read it, and it is remarkably fresh and relevant - a fate few futurist books can claim fifteen years after publications. It’s got great bits on all sorts of things. Gerontocracy. The crudity and just plain boneheadedness of what we think of as “modern” medicine. Love. Art. Transformation. I teared up reading the description of the emotion of holy fire as something Mia, the main character, sees as she looks at a young girl, and how it was an emotion she felt on looking at her daughter. It’s what I feel when I see my son raise his arms on our little patio, water coursing over him from a gardening bucket. It’s a fantastic book. Since I work on privacy these days I’m completely obsessed by the novel’s approach towards privacy. Everyone’s medical information is online to all. Everything, to everybody. Sterling posits a generation burned by a decade of plagues, preferring total transparency to the alternatives. And it’s very plausible. However, it’s not an all or nothing world. It gives us perhaps the idea that what we need is not a monolithic approach to privacy. Maybe we need a new epistemics of privacy - where not all data is the same, and not all things need be shared. There’s a little exchange in the front of the book that sums it up perfectly. “I’m not asking you what networks they’re accessing,” Mia explained. “I just want to know a little about their personal lives.” “Oh, okay, no problem,” Stuart said, relieved. Heh. I’m going back and re-reading Distraction next. As I recall there’s some lovely near-future biohacking in it, as well as great Louisiana dystopia-fiction. Pick up a copy at Amazon.