The Atomic Challenge in Next Generation Education

One of my areas of deep personal passion as a technology enthusiast is education.  I cannot remember who first introduced me to the following anecdote, but it serves to capture the "potential energy" of the space.  If you walk into most classrooms today even at the university level, they look strikingly similar to the way that they would have looked 50 to 100 years ago.  Blackboards, chalk, pencils, notebooks, and - ok - calculators.  This is unlike almost any other industry today.  Think of walking into a commodities exchange or a hospital along a similar timescale.  With the information revolution finding its way into everything from our cars to our kitchens, the lack of relative progress in learning & instruction boggles the mind.

I am proud to have led and been a part of organizations that have spearheaded investments in the sector over the past few years.  For each of the businesses that I have backed, there are hundreds of others that I have looked at - some of which I wish that I had.  The most important transformation that I have witnessed is the migration of pedagogy from a teacher- to a learner-centric model.  Just as the transition of the album- to track-based economies of the music industry allowed consumers to purchase more of only what the want and web services like Twitter, Instapaper, and Boxee give their users the ability to set their own table of contents to the stories that they want to follow, students now have more power than ever before.  

Education for hundreds of years was a market that functioned under local monopolies.  If I wanted to go to college in the pre-plane/train era as a Chicago-an, only the most wealthy of families could afford to send me somewhere beyond Ann Arbor, MI or Champagne, IL.  Now international learners represent a non-trivial proportion of many domestic student bodies.  When combined with the amazing economics of software & the internet, working parents can take supplementary MBA classes late at night or early in the morning from the University of Phoenix, and a child can take chinese lessons from Rosetta Stone if their school does not offer it.   These newer models challenge local monopolies by empowering individuals with tools to achieve a better temporal & geographic fit for their needs.  This is where we stand today.  A working student growing up in the Bronx might just as easily encounter their next door neighbor in an early morning GMAT test prep course as a Indian accountant burning the midnight oil.

The learners of today, however, demand more.  In an era where the music streams of Pandora adapt to each listeners love of lyrics and temperament toward tempo, educational technology must find ways of teaching the same routine concepts math, science, language arts, and social science in ways that cater to the specific learning needs & preferences of the individual.  Some students learn geometry better through equations.  Others grasp it better through graphs.  Some students learn language better via video visualizations.  Others keep up better when forced to spend time conversing in groups.  The only way that the educational system can scalably be more finely tuned to optimize pedagogical concepts at the learner level is through technology - an exciting opportunity for the entrepreneurial ecosystem.

This is fundamentally a data problem, one I call the Atomic Challenge in Next Generation Education.  The only way to "optimize" is to know by student what techniques and content are working and what aren't.  The only way to do that is to tag each and every question posed by difficulty, format (graph-based, video, algorithmic...), time of delivery, and really any other variable that could possibly effect an individual's performance.  Next, a level of machine learning is needed to build a customized content feed according to the formats that maximize that individual's chances of success.  

Atomic data-based learning is going to be the way of the world because the failures of the system - at least in the US - demand it.  With the staggering cost of healthcare and impending bankruptcy of the Social Security Trust, their economy must grow fast enough over the long-run to afford to pay both principal & interest to reduce the national debt.  Super-charged & sustainable growth requires heightened levels of productivity.  Since the 1980s, however, the US has experienced half the productivity increase as in the previous two decades.   The principle way in which we boost productivity is by fixing our falling educational system.  In order to improve, we need to understand what is working.  And, in order to understand what is working, we need data.  The Duncan / Obama $4.35B "Race to the Top" program embraces "Standards & Assessments"-based initiatives as a key area of focus.  Only by accepting that reality will projects be eligible for state financing.  When teachers fail to improve the performance of their students and cannot get our young to basic levels of literacy, we need to know.  

Along with the district & state bureaucracy that supports them, institutions of instruction have shifted to an Age of Compliance.  Accountability will soon pervade schools the way it governs our hospitals and courts.

If we take for a given that today's relatively data-lite schools will some day be guided by 1s and 0s, what sorts of technologies exist to enable transition?  How can we think about building educational content in the absence of a common set of pedagogical formatting standards?  Is it worth it, or should we wait?  These are questions that keep me up at night.  If you have ideas, drop me a line.