The Naked Future (18 page)

So far, the system's biggest asset has been a collapse in the cost to do certain types of education and curriculum experimentation. “In a traditional education study, you may have twenty students in your experiment group and twenty students in your control group. And if you're really lucky, maybe you get something that's just barely statistically significant,” says Ng. An extremely large education study, encompassing on the order of hundreds of students, can cost thousands of dollars and sometimes won't produce actionable results for months or years. On Coursera, every interaction can become an A/B test in which one-half of the test-taking population is shown one lesson and the other half is shown a different one. On any given day Ng can run such a test on twenty thousand students. All of this data is helping him understand the process of learning in a way that is specific to any individual and yet broad enough to be applicable to any student in any country.

In 2012, when two thousand of his students submitted the same wrong answer to a question, Ng realized immediately the problem wasn't the kids taking the course. Rather, the malfunction lay in the question itself: “In a normal class, if two students out of a hundred submit the same wrong answer, you probably won't even notice. But when two thousand out of a hundred thousand students submit the same wrong answer, that's a very strong signal to the instructor that some clarification is needed.” Ng checked the question again and at first didn't see any error. Most of the students got
it right. But when he applied a learning algorithm to the data set, he discovered that the wrong students were all making the same type of error: they were reversing two steps in the formula. In effect, he learned to predict errors before they occurred. He created a customized error message so now every student who misses that question is given a clue, a message urging him or her to go back and reconsider the order of the operation, which allows the students to correct missteps and move on much more quickly.

These sorts of discoveries are increasingly common in online environments that actually collect and use student data. When MIT physicist David Pritchard first analyzed the results of a widely used concept test given to about a thousand students, he found that most of the carefully designed problems tended to challenge the problem students but were a breeze for the better-prepared kids in the class, as anyone might have expected. But two problems in particular produced a counterintuitive result. The A students did well and the B students did less well; however, the C and especially the D students did better than the B students! He realized that the wording of the question was ambiguous, but the A students seemed to know what was being asked. Many of the students doing less well, after misunderstanding the question, had a common physical misconception that resulted in the correct answer—two errors combined had a canceling effect. It's the sort of event that a regular teacher lecturing to a class of thirty would never notice but because Pritchard had hundreds taking the concept test, the size of the data set enabled analysis that made the error visible.

Rapidly adjusting lessons on the basis of new incoming information is only possible in the Internet era. But Ng is careful to point out that he's not actually an advocate of Web-only education. He's a fan of what's sometimes called the flip model, which is in-person education with a heavy online component.

Far more students today have seen Ng's lectures because of the popularity of the machine-learning course, but he actually spends much less time lecturing than he used to. He uses his class time to teach high-level concepts, try out new material, and workshop big
problems. “The instructor can look at a Web site and see what the students are getting right and getting wrong, so that [the instructor] can focus the classroom discussion on what the students are actually confused about.”

Some of the most pioneering work in this field was conducted by Kansas State University physicists Dean Zollman and Sanjay Rebello
,
who handed PDAs to their students in 2005 and instructed them to actually text in class (so long as the subject was physics). Zollman and Rebello were able to quiz every student in real time and then alter lesson plans accordingly.

In many ways the Rebello-Zollman classroom provided a snapshot of what Ng is doing, and, perhaps, all classrooms of the future. In a press release Rebello remarked that the system worked well to address issues of minority student engagement. “I find that even in a small class it can give me a feeling for how this silent majority or silent minority of students is thinking about things that I wouldn't [normally] get.”

Ng already has a lot of competition in the online learning space. On the East Coast the edX program, which features online interactive courses from MIT and Ivy League schools, has already attracted thousands of participants. On the West Coast, two of Ng's colleagues at Stanford, Sebastian Thrun and Peter Norvig, put their own interactive course in artificial intelligence online at around the same time and are now spearheading a company called Udacity.

Norvig also serves as a director of research at Google and is one of the most senior executives of the demographically young company. Prior to joining the search giant he led the computer science division at the NASA Ames Research Center where he worked on sending robots into space. He's considered one of the world's top minds in AI but is also known for his oddball sense of style. If you catch Norvig at a big event, he'll likely be in one of his trademark Hawaiian shirts. In public he speaks in a slow and swampy-deep voice that seems to emanate from the very bottom of his six-foot-and-then-some frame. The mathematical genius and the inner joker seem to have reached a strange but solid equilibrium. In teaching his
online course, he found that little mistakes and moments of goofiness that can screw up a live lecture actually work well online.

“I thought it was about recording videos and making them flawless. What I discovered is that it's about making a personal connection. Finding a way for the students to make a commitment to do the work,” he told me.

I first met Norvig at the 2007 Singularity Summit in San Francisco. We chatted for a bit and the subject soon turned to education. At the time I was needlessly concerned about the effect that Google could have on the future of learning. I pointed out that 50 percent of high school seniors (at the time) couldn't tell the difference between an objective Web site and a biased source. I asked him what he would do to preserve critical thinking skills in an era when such technologies of convenience as Google seemed to be doing a lot more “thinking” and students a lot less.

He answered, “What can you do about that? I think part of it is education. We're used to teaching reading, writing, and arithmetic; now we should be teaching these evaluation skills in school and so on. Some of it could be just-in-time education. Search engines themselves should be providing clues for this.”

When I ran into him again at the 2012 Singularity Summit he was about ten pounds lighter and still wearing his trademark Hawaiian shirt. He was well on his way toward fulfilling the forecast he had made several years earlier: using the technologies of data storage, retrieval, and machine learning to bring just-in-time education closer to reality for millions of students.

“We're going to make rapid advances in understanding what works and what doesn't on the basis of interaction statistics,” he told me. I asked him if he believed every student would have access to a predictive model of their own learning style in the next five years; if, in effect, there was a naked future for education. “That's the hope,” he answered.

Today, online learning platforms that measure individual student performance are in their infancy. Andrew Ng, Sebastian Thrun, and Peter Norvig are still learning what works and what doesn't. But
it is within the capability of the Coursera platform to offer individualized instruction on a level that surpasses what almost any in-person teacher today provides.

“We are analyzing the data as we try to understand when students are likely to have problems, when they are at risk for dropping out of the course. These are things we're looking at. And eventually, hopefully, this will allow us to catch and to encourage the students along when they need that bit of the extra encouragement. These are things we're working on. It hasn't happened yet,” says Ng.

So far in this book I've covered some of the ways we've begun to accept rapidly evolving notions of privacy. In such areas as epidemiology, sharing more personal data can have a positive effect but at the personal cost of telling the public that you're sick. It's nakedness in the worst sense: the benefits are public and the costs are private. Learning and education is one example of where individuals will begin to see benefits of exposing a bit more of their life stream. Those who elect to take a different course will find themselves ever more at a disadvantage.

We haven't begun to understand how those depreciating costs of collecting and analyzing real-time scores, footnotes, and time-stamped marginalia will change the definitions of teacher, student, and learning. If Andrew Ng and Peter Norvig's experience is any indication, teachers are going to realize the dual and compounding benefits of being able to provide much more personal attention to a far greater number of students. The teachers who are extremely effective at this will model the way forward for everyone else. But in the coming decade, teaching will come to mean something very different than what it means today.

It's January 2012, the setting is the Ethiopian village of Wenchi, which sits on the rim of a volcanic crater lake some eleven thousand feet above sea level. Wenchi village is one of the poorest communities in
one of the poorest countries in the world. The majority of the houses here, without running water or electricity, are scarcely larger than a single room. They have dirt floors and roofs constructed of branches, which seem to provide little more than shade.

A group of non-governmental organization (NGO) workers arrives in the village from the nearby capital of Addis Ababa. They bring with them a sealed cardboard box. The workers have a brief conversation with one of the village elders and explain that inside the box are tablet computers, one for every child in the village ages four to eleven. The tablets have been outfitted with solar screens so they can charge in the sun and come fully loaded with hundreds of apps, movies, and games, all in English. English is the official language of this village and this was the chief criterion in its selection for this gift. But “English speaking” is just a technical designation because the population of this village is illiterate down to the last person and unable to understand instructions or subtitles that are part of the apps or the movies on the devices. In fact, almost no one here has ever encountered the written word in any form. There are no street signs, no candy wrappers, Coke bottles, flyers, or advertisements. It is a tabula rasa in the desert.

The relief workers from Addis Ababa leave the box and return to the capital. They will circle back to the village soon, in about a week or so, to swap out the subscriber identity module (SIM) cards in the tablet PCs.

Software running on the tablets will log every keystroke and swipe when it occurs. The devices will record each child's progression through the games and apps, step-by-step, command by command. The cards will then be express mailed from Addis Ababa to Cambridge, Massachusetts, where Nicholas Negroponte and a team of researchers he's assembled will analyze the data and search for clues to reveal how the children are improving at written and spoken English despite having no instruction whatsoever.

Skip ahead a few months. On October 25, 2012, Negroponte stands in front of a gathering of technology enthusiasts at MIT to relay what months of research have uncovered. “In five days they
were using forty-seven apps per day,” says Negroponte. “In two weeks they were singing ABC songs.” By the end of five weeks, Negroponte reports that several were able to activate the tablet's disabled camera and take their photos. “They hacked Android!” he shouts to the crowd.
³

Negroponte is the founder of the MIT Media Lab. He's credited with being the first investor in
Wired
magazine and is the author of the 1995 bestseller
Being Digital
,
a forward-looking treatise on the future of man and machine that seems to grow only more influential with the spread of the Internet. He's probably best known for his One Laptop per Child (OLPC) Association, which donates specially designed laptops to children across the developing world.

When you go to the association's Web site, you'll see pictures taken of OLPC initiatives across the globe. One in particular has become closely associated with the program. It features several dozen Nigerian boys all about age ten. Each one is wearing a school uniform of avocado green with forest green trim. They sit in neat rows at long schoolhouse desks. Each boy has a bright white and green laptop. These devices, set against the blandness of the cement wall classroom, the rustic benchlike tables, the dour green uniforms, look like artifacts from the future deposited in the past. All of the boys are smiling politely.

A casual observer would say the picture speaks to the enormous success of the program, which has given away more than 2 million laptops around the world since the first full model was revealed in 2005. But Negroponte admits to a degree of ambivalence about this scenario.

“This was the image that started to spread because, sadly, our partner was always the government, and the government wanted us to do schools.” He says “schools” with a distaste that's reminiscent of a ten-year-old at the end of summer vacation.

Negroponte doesn't measure the success of the program on the basis of units distributed or test score improvements. Not only are these the wrong metrics but he believes the continuous focus on
testing, curriculum, indeed on
teaching
, often gets in the way of real
learning.
The assumption that the latter is impossible without the former is wrong. “We know that a vast recall of facts about something is in no way a measure of understanding them,” Negroponte wrote for the online edition of
MIT
Technology Review
just before his lecture. “At best, it is necessary but not sufficient. And yet we subject our kids to memorizing. We seem to believe that rote learning is akin to physical exercise, good for their minds. And, quite conveniently, we can test whether the facts stuck, like spaghetti to a wall. In some cases knowledge is so drilled in that you know and hate a subject at the same time.”
⁴
For Negroponte most of what we call school is a gratuitous time suck at best, and a real impediment to learning at worst.