Home » Blog » ideas » Tim O’Reilly at MIT Media Lab

Tim O’Reilly at MIT Media Lab

Legendary technology publisher Tim O’Reilly is the speaker at today’s Media Lab Conversations with Joi Ito. O’Reilly Media is one of the best respected technology publishers in the industry, producing many of the books working programmers rely on to build their software and systems. O’Reilly is also an important convener – the conferences his organization hosts are important in shaping the dialog about the internet and innovation.

Tim opens his talk with two quotes, one from Oscar Wilde (“Quotation is a serviceable substitution for wit”) and a second from designer Edwin Schlossberg: “The skill of writing is to create a context in which others can think.” Tim suggests that where he has been most influential and powerful is around presenting big ideas that create context for how others fit. In practice, he publishes books and organizes conferences – in a larger sense, he’s constructing ideas and frames.

He considers his first big success to be changing the framing around open source software. There had been a dialog around “free software” from Richard Stallman, which Eric Raymond helped push towards a discussions around “open source”. But Tim worried that these guys were just talking about Linux, and wondered why they weren’t talking about the web or about DNS. He realized that Stallman’s argument was primarily a political one, and that Eric had moved the narrative slightly, but fel that neither was creating a broader context that showed what the projects had in common. The internet itself is an example of open source software and open source thinking.

A second success came from developing Dale Dougherty‘s term “Web 2.0” The term was coined to make the case that, despite a crash in the tech market, the web wasn’t over, but would come back. Tim expanded the idea into talking about the web as an operating system, supporting subsystems like mapping and identity which can reside somewhere out there in the cloud. The specific term – Web 2.0 – doesn’t matter, but the shift in framing does.

Tim cites Dale Dougherty’s work with the Maker movement as another form of shaping big ideas, connecting ideas that people previously saw as disparate into the same frame. The first Maker Faire featured Swapper Rama Rama, a group that recycled clothes by sewing them into new patterns and held a fashion show at the end of their session, and the Alameda/Contra Costa computer recycling project, which featured a biodiesel-powered linux supercomputer made of recycled PCs. Prior to the Maker frame, no one had seen these projects as interrelated.

Recently, Tim’s activism around open government has been focused on trying to make another conceptual shift, from seeing government as a platform, rather than as a solution provider. We spend a great deal on government, Tim tells us, and it’s often not serving people sufficiently. We want government to be like our phones – we’d like to spend less and see more benefits, much as we went from carrying phones with a few dozen applications to a current world where we can choose from hundreds of thousands. If government is a platform, as the iPhone or Android are platforms, can we expect people to build hundreds of thousands of potentially useful applications?

The process of framing big ideas can help people see what’s in front of them. We need to look with “soft eyes”, Tim tells us, considering what we see until a pattern begins to emerge. He offers two ideas that, for him, are still in the soft eyes phase:

One is the idea that we’re building a global brain. He suggests that the companies that survived the Web 1.0 bust were those that leveraged collective intelligence. Google uses the signal of clicks to tune search algorithms, Blogger uses people’s participation to create content, Twitter creates value from helping people transmit information around the globe.

He quotes Danny Hillis – “Global consciousness is that thing that decided that decaf coffee pots should be orange” – to explain ways in which knowledge can emerge from behavior. Twitter works like this, he suggests, when hashtags emerge. And this is now how we’re learning to program driverless cars – the progress from an autonomous car that, in 2005, drove seven miles in seven hours to Google’s cars that have driven thousands of miles, is based in no small part on collecting data from human drivers. Building Streetview has allowed Google to collect a vast set of data about how humans navigate streets all over the world – a robot car can “remember” human driving experience on existing streets.

We might think of these new systems as a form of man/machine symbiosis, where augmented humans and connected computers work together, helping computers do things they couldn’t do before. The idea comes from J.C.R. Licklider, who anticipated much of modern computing. Our challenge is to engage in this sort of symbiosis at scale and at speed, not necessarily realtime, but realtime enough. He reminds us of a question asked by an IBM executive: “Would you cross the street with information that was only five minutes old?” Of course we wouldn’t – but we need to figure out exactly what we’re optimizing for and how real-time is enough.

Another way to think about this symbiosis is to realize that the programmers are always embedded in the code they create. Tim tells us that he spoke to Amazon about the idea of the mechanical turk well before their product of the same name – for Tim, the mechanical turk, the man within the machine, was the way in which the developers of Amazon’s software lived on within their systems.

Tim’s second, more inchoate idea is offered up as a possible illustration of how his mind works: he enjoys probing on soft spots, on areas where it seems something is not quite right. In discussions of SOPA/PIPA, Tim has noticed that there’s lots of quantification of the cost of piracy, but very little narrative about quantifying the value fo the internet.

Tim tells us that he pays $60 to his cable company for HBO, and that we all understand this as a content transaction. But he also pays $60 to his cable company for Facebook and Twitter, and there the content is enabled by companies, provided by individuals. It’s free content… and there’s something slightly wrong in that narrative.

He references a paper in Stewart Brand’s journal, the Coevolution Quarterly titled The Clothesline Paradox. If you put your clothes in a clothes dryer, the energy you spend is part of the measured economy. Hang them on the clothesline and they disappear from our economy. How much of internet value we are creating isn’t being captured or measured?

It’s fine not to capture all the value possible – Tim suggests that his corporate motto is “Create more value than you capture”. He references an unnamed internet billionaire who told him, “It all came from one of your books.” Tim notes, “Hey, you got billions and I got $35” But that’s okay. The titans of Wall Street have been great at creating wealth for themselves, while Tim Berners Lee created immense value with the World Wide Web and captured almost none at all.

We are starting to see thinkers like the current CEO of Bluehost who want to find ways to give something back to the Open Source community to recognize how much value they’ve gotten from the software ecosystem. Bluehost’s CEO explained that he’s got almost 20 million small business customers, and that they’re able to serve them due to open source software. He cited a McKinsey study that companies with websites have 23% more value than those without – for Tim, that might be a shortcut to measuring and valuing the power of the internet.

“What are the hidden economies we should be measuring to consider value within the internet ecosystem?” We might think of this as tying into the global brain discussion, Tim tell us. If we really are building networks of collective intelligence, can we capture value from those networks? We’re all contributing to google or twitter – in some ways, we get that value back. Is that value sufficient? Should Google pay more to YouTube creators for the videos they make? Can we build a more robust economy building on ideas like Cory Doctorow’s, that our attention has value and that we should consider monetizing it?

Joi begins his questioning of Tim by asking about this insistence on economic value. Yochai Benkler helpfully points out that we often collaborate because we like to share. Clay Shirky explains that we’re working through our cognitive surplus. Is economic value really the way we should consider web value? Tim concedes that he’s in the early stages of this thinking – economic value seems to be the logic that holds sway in Washington, and we may need to learn to speak this way to influence those debates.

Reflecting on Tim’s discussion of a global brain, Joi asks what Tim thinks of singularity theory. Tim cites Samuel Johnson’s refutation of Bishop Berkley’s ideas about immateriality: kicking a stone and announcing “I refute him thus.” It’s one thing to predict disembodied intelligence from extrapolative graphs, but something entirely different to demonstrate it. Tim reminds us that, for a thousand years, the Hagia Sofia in Constantinople was the largest building in the world. And then the Blue Mosque, built across the street, was the biggest building, a near literal copy. It’s possible for human progress to stall for long periods of time. We might face a stall through exhaustion of fossil fuels or political calamity… and once progress slows, it can be hard to get back on the horse.

At the same time, he allows, the notion of singularity at least posits a worthwhile and ambitious challenge. When Tim looks at lists of “tough problems” listed in the physical sciences, he’s struck by how boring and dull most of them are. Thinking about the singularity involves thinking big and broadly.


I was lucky enough to have Tim drop in on our Participatory News class before giving his talk, where I and my students asked him about how O’Reilly Media has thrived in a digital age given the shifting nature of book publishing. He explained that his books now make about 40% on average of the revenue that they made when he began the business, but that as a whole, the business is far more successful. It’s not that the conferences subsidize the book business – the books make money on their own. It’s that they are much more efficient to make – they tend to be printed on demand, the editing process is streamlined… and authors make less than they used to.

Tim suggests that his success has come from balancing two factors – paying attention to what’s interesting to him, and to what people want to spend money on. Were he purely chasing the money, he’d miss much of what was going to be interesting and wouldn’t have the role he’s had in shaping thinking about technology. I pointed out that Tim’s work follows a third factor – what’s consistent with his mission, which centers on technology, knowledge and progress. Efforts like promoting open source weren’t motivated solely by Tim’s interest in the field or a belief it would make money, but by a sense that it was the right thing to do.

It’s possible that Tim’s success has had to do with the way these three factors – mission, interest and revenue – interact. What’s interesting and mission-driven often becomes an area where people want to know more and spend money. Our class spent the time after Tim left wondering whether that three-fold model could work in the journalism space. What happens when journalists believe something is key to the mission – informing a public – but isn’t where the market wants to go? Is it possible that your interest and your sense that something is in support of a mission is sufficient to lead to a market for that story? Is journalism about making that leap of faith, or about a better understanding for what markets want?

5 thoughts on “Tim O’Reilly at MIT Media Lab”

  1. Thanks for posting this recap.

    A few thoughts re the Singularity: I’m a fan of the work Kurzweil has done with his curves, these patterns of exponential growth open some fascinating avenues for thought. In the framing Kurzweil gives the curves seem to have a life of their own, a sort of intentionality or physical law that necessitates exponential growth, that these curves are the equivalent of a snowball that can’t be stopped.
    I love Bill Joy’s counterpoint, an exploration of the potential dark side of the singularity – that we may not be ready for the powers that will come as we climb up Kurzweil’s curves (available here in three parts – http://www.youtube.com/watch?v=kbqdhyz9u8M&feature=relmfu)
    Recently I’ve been thinking about our role as curve shapers, that the choice isn’t a binary on (Kurzweil) / off (Joy) but perhaps were part of a move toward fine-grained, global democratic control over the slope of these curves. Pour more resources/incentives into one area in order to increase the pace of innovation and remove incentives/resources in another in order to slow, stop or even reverse the course of innovation. And there is always the possibility of some exogenous shock, from an asteroid to a plague, that will impose its own direction…
    Whether continued innovation will lead to the singularity or not, there is no doubt that we collectively face a series of grand challenges and that the implications are global. My recent work has focused on how do we make these important global decisions at speed and scale? What new mechanisms of education and action and governance do we need in order to collectively & effectively engage these issues? Surely we’ll need new tools in order to manage the process.
    I’ve written a bit about the issues of curve control and decision making here – http://restartdemocracy.org/2011/10/07/we-changed-the-slope/

  2. Pingback: Tim O'Reilly on Internet privacy and the promise of Web 2.0 - Sally Duros

  3. Pingback: HuonFM Tech News and Blues 2 Jul 2012 « David Coldrick’s Blog

  4. Pingback: Government as a platform rather than as a solution provider by Tim O’Reilly | The Future of the Net | Scoop.it

  5. Pingback: Tim O’Reilly at MIT Media Lab | Lukor.net

Comments are closed.