Tuesday, April 29, 2014

Hacking University Math: Yes We Khan!

Hacks -- the successful ones anyway -- typically mashup existing technologies in innovative but attainable ways to produce real change. So I'm taking the opportunity of presenting my education hack to you all here at FutureTense:Hacking the University Ed very seriously. It's a golden opportunity for me -- say my hack were to match the resources and interest of enough of you here today to make it happen! That would be change we could believe in -- a much faster change than we've come to expect from the academy at scale.

My colleague Johnny McCoy says it pretty well here, I think. Our academic institutions oftentimes feel maddeningly slow to change. Our schools do so many things so well that we have a natural concern to avoid throwing the baby out with the bath water. Take that reasonable concern and combine it with the interlocking web of interests and traditions that seem to cement our current practices in place, and it can be hard to see how and when real change could ever come.

This patient, forgiving-looking man is Michael Crow, the president of Arizona State University, architect of the New American University. At last weeks EdTech summit in Scottsdale, President Crow asserted that the principal obstacle to obtaining the significant improvements to educational outcomes that we are all seeking is CULTURE CHANGE. The technologies are here, the investment is here, the desire is here. What's to often missing are the levers needed to change the culture.

An example will serve to prove the point. Consider the great store that educational reformers, myself included, put in the potential of adaptive technologies to personalize education. Adaptive, personalized education is one of those things it's hard to be against. Big data and other machine learning techniques are producing such wonders (Watson, Google Predictive Search, etc) that proponents expect adaptivity to dramatically change the effectiveness of education -- by providing teachers and students with unprecedented insight into student understanding over time and making truly personalized learning paths practical. Fans like me not only point to the effectiveness of today's technology, but expect great things to come, as these technologies evolve over the next several years -- as databases of student learning grow; as BigData and other machine learning approaches advance; as human experience with their effective application grows over time. Not everyone agrees, but among those who do much is expected.

Which is why Math is such an important bellweather. If adaptive technologies are going to live up to their promise and make substantial improvement possible, we will see that progress first in Math -- before all other subjects. Math is such an important human capability, the gateway to deep understanding of science, engineering, and computing. Math is the very basis of rational thinking, the one discipline that all the peoples of the earth agree upon, precisely, to the very last letter. Its the language that computers speak natively. So much more cut-and-dried than other areas of human activity, Math is uniquely suited to machine evaluation and machine guidance. If computer tutors are going to help us learn anything -- we will see the evidence in Math first.So, when's that going to happen?

I think it may have already happened. Education may already have seen its first Napster moment, that moment when a new technology blows open an established practice by operating at a new level of scale -- think of the rise of Wikipedia. In August of 2013, the Khan Academy released a new capability. Moving beyond the library of low-production value, high explanatory-value videos that the Khan Academy use to explain most of math (and several other subjects besides) to millions of learners, the Khan team released a basic adaptive math tutor -- operating at Internet scale, offered to the world, for free. Since its release just 8 months ago, already millions of students have used the Khan Academy to improve their performance in mathematics. Many have used it in school, but many, many more have used it on their own, as a supplement to the other ways they have of learning math. These students come from all over the world -- from all over the internet, anyway. The Khan Academy tools are available to anyone with an Internet connected device -- say an ipad, a Chromebook, a web browser on a laptop, or a phone. That's all anyone on Earth needs to get access to this adaptive computer tutor.

The tutor poses problem after problem, exploring the depth of your understanding -- incrementally challenging you along your own path. It patiently watches you do every problem, keeps track of how many attempts you make, how many you times you get one right, how many times you get one wrong, how often you say "I don't know", or "This is too easy", how often you ask for a hint, and how often those hints help you get the right answer. Gradually, student by student, problem by problem, day by day, the Khan Learning Dashboard is itself learning, amassing the largest, most detailed, most valuable database of human learning behavior ever assembled. And they're just getting started...

The Khan Learning Dashboard understands the connections between the concepts in math. It is aware of the pre-requisite dependencies that govern the feasible paths to mathematical understanding. In addition to using those relationships to create a personalized learning path for each student, the Khan Learning Dahsboard provides detailed progress information to students and their coaches, providing learners and teachers with a more complete picture of each student's understanding than traditional tests and homework can provide.

The Khan Academy is on its way to becoming one of the great Internet Utilities -- one in a family of unbelievably valuable, and often technically complex and expensive tools -- that are provided to all the citizens of the Internet for free. Tools like Wikipedia, GoogleSearch, Twitter, Gmail, Google Docs, YouTube, Facebook that are so useful, so valuable, and yet so free!!

Based on a donation model and backed by some of the largest educational foundations in the world -- the Bill and Melinda Gates Foundation first among them -- the Khan Academy is uniquely positioned to provide this core Internet utility at no marginal cost -- making it a friction-free choice for educators at every scale, every grade, around the world. If Khan's team continues to execute, there's no reason the Khan Academy can't stand right next to Wikipedia on the world's electronic bookshelf as one of the greatest educational resources of all time.

It's important to understand Khan as an Internet Utility, because it's important to consider not only what the Dashboard is capable of today, but also how rapidly it can be expected to evolve year over year. It's not just about what the LearningDashboard can do today; it's about how it will increase in capability as API's are established, as the database grows, as the machine intelligence learns, and as the human pedagogical practices that surround its use expand and improve through shared experience.

I can't say how much better Math teaching will get. I can't tell you what the technologies that ultimately grow from these roots will be like, or exactly what they will be capable of. But I expect them to get a lot better, and quite quickly. Who expected Assasin's Creed when they first saw Pong? But if the exact form of progress is unpredictable, the magnitude of that progress seems completely predictable.

So, you ask, "What's preventing educators of all stripes from using the Khan Academy with their students, helping shape the development of this resource and contribute to the development of effective pedagogies that incorporate Khan's rapidly emerging, very promising tool?"The short answer is -- a fundamental change in the academic culture; a change in approach that acknowledges that explanations of complex concepts are no longer scarce, but plentiful -- moving a math teacher's primary locus of value away from being a lecturer and towards being a guide. Our experience at ASU tells us it is a profound cultural shift, but a rewarding one; a change with both room and promise for improvement.

"I'm sold", you say. "If only there were a vector for bringing the necessary cultural change about, to organically spread best practices, in a grassroots way; to form a coalition of the interested, the willing, and the able; a way to establish fruitful partnerships between Foundations and Universities to sponsor and support teams of educators, administrators and technical professionals, who come together to use the Khan Academy to help more students succeed in math -- and in the process collect, disseminate, and foster the improvement of those practices that prove to lead to enhanced student success."

Well, I have good news! There is just such a model, and copying it is the foundation of my educational hack. The model is the "Khan Academy in Idaho" initiative sponsored by the J.A and Kathryn Albertson Foundation. My hack is really simple -- that's what makes it feasible. It is to replicate the "Khan Academy in Idaho" model in every state in the Union -- Khan in Connecticut, Khan in Kansas, Khan in Colorado, Khan in Arizona. To combine, in every state, a local university with a strong commitment to computer-aided mathematics teaching and a strong tradition of outreach to k-12 teachers with a sponsoring Foundation, committed to making a meaningful contribution to the state's educational improvement at scale. The federated result will be a national guild of educators who have learned how to effectively use the growing body of machine intelligence to help more students succeed at mathematics; a guild, who will introduce a personalized, adaptive education culture into their local institutions.

We've seen higher education's technical culture changed by the introduction and rapid adoption of an emerging Internet Utility. Not so long ago, a university choosing a free, cloud-based email utility to replace their homegrown system was considered a risky move. Now it's de rigeur. It was a complicated cultural change and it happened much more quickly than most expected.

These Internet Utility models scale like nobody's business. All the investment happens at the center, in a concentrated way. So Internet utilities can fully exploit the advantages of cloud infrastructures; they can improve very rapidly and spread very fast. Cultures can change around the use of these public utilities, because all the users benefit from the central investments. For a thing like a personalized math tutor, the approach is perfect. More users brings more data. More data brings deeper understanding. Deeper understanding leads to better outcomes. Better outcomes bring more users. Wash. Rinse. Repeat.

The Khan in X "change model" relies on individual human decisions, made at various levels in the hierarchy -- but always from the bottom up. Not boards, corporations, lobbyists or legislatures, but Teachers, Principals, Superintendants and Technologists --- each deciding, within their own spheres of influence, to explore an approach that has worked elsewhere, in a situation that looks to them a lot like their own. A local foundation and a university can quickly form the nucleus around which a coalition of the interested, willing and able can assemble -- at nearly any level of math education. The "Khan Academy in Idaho" model lays out the basic blueprint. An online federation of "Khan in X" initiatives will make it possible to rapidly recognize and share a wide array of best practices -- old and new -- that can be shown to work to foster the success of an ever broader spectrum of learners.

So now is what we call in the pitching business the "altar call", where we find out if my pitch was compelling enough to get you to take action. To make it less embarrassing for all of us, let's just do it over Twitter. If you thought this pitch was worth ignoring, then just continue to ignore it. But if you were genuinely interested, please consider tweeting about it to your followers: maybe use the hashtag @sannier for criticism and #khanInIdaho for praise. If you're interested in learning how to help start a "KhanInX" chapter in your state, tweet me at @sannier #khanInIdaho. Follow those keywords and you can see how flat this pitch fell :-)