Why a Publishing API? or What's a Publishing API?

The fellowship that I have through the Shuttleworth Foundation is officially “to foster an ecosystem around open education resources”.  The community creates great teaching and learning resources, and then the ecosystem makes it possible for educators, innovators, students, and life-long learners to use the resources, improve them, and connect them together in novel ways that enhance learning.

When you look at the details of my fellowship proposal, the first thing that I am starting with is creating a publishing API (application programming interface). The API part of that means that it is a language that software programs speak to each other. If you use TweetDeck to update your Twitter and Facebook accounts, then you are using software programs that are talking to each other using APIs. (Because Twitter provided a simple API early on, tons of services I had never heard of use the Twitter API. A cooler-than-me friend sent me this list of programs that use Twitter’s API -- “Brizzly, Seesmic, Tweetie, DestroyTwitter, TwitterFox, SoBees, Mixero”.  And another mentioned that all of our email programs communicate with each other via a lovely open API called SMTP. Email succeeded because it is easy for programs all over the world to talk to each other and deliver your email. In fact, the web and the way browsers display web pages are all built around API’s.

What do I mean by “publishing”? Here, I am talking about creating educational resources and then making them available to others in a library or repository. Making them available is the publishing part. So if you create a nice lesson for teaching fractions to middle school students and then want to share that lesson, you might publish it by uploading it to Connexions (cnx.org), or a Google Doc you make public, or your school’s website or Open Courseware (OCW) site. And then you might post a link to it on Facebook and Twitter and email the link to colleagues. Some of those colleagues might want to download it, add an exercise that they have used successfully, and share that updated version.

Now imagine that I have an idea for a simple editor that also creates really nice equations and that I want to build in a “publish” button that will let you choose where your content should be published, figure out what format it has to be in and convert it if need be, and will also advertise your content to your social networks.  If places to share the content (like Connexions) support a publishing API, then my editor can do all that for you. This scenario is only one example. I intend to blog many more examples from conversations I am having with other people and projects that are building software and content for education. These are the sorts of possibilities that a publishing API will support. I am also going to do a few more technical blogs on what things I think should be in that publishing API and what other APIs already do much of the work and provide a good place to start.

Combining the best of the studio model with personalized learning: Is it doable?

I was recently at the NITLE (National Institute for Technology in Liberal Education) Summit meeting in Arlington, Virginia and John Seely Brown (JSB) gave the keynote address on “A New Culture of Learning for a World of Constant Change.” His speech made me wonder whether new virtual learning environments can combine the benefits of the studio model and the benefits of automated, but highly-personalized learning. But first a little summary of his talk.

First, he proposed three fundamental shifts that define the world as we know it in the 21st century.

1. Explosion of data -- Google’s former CEO, Eric Schmidt, says that every 2 days we create the same amount of information as we created from the dawn of civilization to 2003. I found an interesting analysis of this quote, but regardless of the exact numbers, the general trend is very clear.
2. Exponential advances in computation storage and bandwidth: These shifts have led to cloud computing, GPU’s (graphics processors), machine learning that automatically processes vast amounts of content and usage patterns.
3. Large-scale, deeply-connected problems. Grand challenges require an interdisciplinary, socio-technical (human process and technology) approach. As solutions are implemented, they change the problem.

The result of these changes are that the half life of skills is shrinking dramatically.

So the question that JSB poses is how do we educate people that will be able to thrive in this environment of constant change, discover opportunities, and tackle the grand challenges?

His idea, as I see it, is that the goal is to see ourselves as “designers”, “creators”, “producers”, and “makers” and to have the ability to empathize with others so solutions will be usable. People will constantly incorporate new skills in the pursuit of the current challenge and that will seem natural, rather than overwhelming. The ingredients for the human as designer are knowledge, play, and, making.

The studio model, where individuals or teams share a physical space and work in parallel on similar, but unique projects, provides the ingredients for learning and absorbing the identity of “creator”. In a studio, experienced “masters” provide critiques (advice that moves a project forward along its own trajectory), and everyone critiques, is critiqued, and benefits from the critiquing of other projects. Essentially, his recommendation is to incorporate this model into education as much as possible. (If you don’t know Olin College of Engineering, definitely check them out. I recently saw a talk by one of their faculty and they have embraced the idea of the studio model of learning completely. One of their design challenges has 5th graders judging the swimming ability and aesthetic appeal of college students' robotic creations.)

My questions: I have been interested in “learning machines” research; investigating personalized and optimized individual learning that takes advantage of the data, storage, and computation now available to deliver knowledge and practice to students, just-in-time. Another benefit of open education resources (OER) is a giant pool of content for feeding into learning machines, thus tying in my fellowship goals.

If the studio model covers “making”, could learning machines cover the “knowing” part of the ingredients for human as maker?

Can the studio model be virtual without losing effectiveness, and maybe even create gains? Another Shuttleworth Foundation fellow, Philip Schmidt, co-founded P2PU, the Peer 2 Peer University, where learners organize courses and deliver and take them together, virtually. Others are also creating virtual environments for learning (University of the People, OpenStudy, Khan Academy to name a few). How much should virtual environments try and mimic the real world where people gather in one space at one time? Are critiques richer if they are delivered synchronously? Would thinking explicitly about how to incorporate more from the studio model enrich these environments?