Teaching Machines

Technology offers real possibilities for positively changing the way we relate to each other as human beings. For example, adaptive technology for children with special needs gives us the potential to integrate even children with severe disabilities into general-education classrooms.

But one laptop per child can’t lift communities out of poverty, because technology is not an alternative to wealth redistribution from the top 1 percent to the bottom 99. There is a disconnect between what we imagine technology and education can do, and what they actually do.

Management gurus and their tech-industry followers insist that if we can dream it, we can do it; that instead of “throwing more money at the problem,” we must use our creativity to brainstorm best practices for education and make them scalable. Harvard Business School Professor Clayton Christensen believes that in the future, computer-based instruction will entirely replace the current model, bringing a higher return on investment for the nation’s education system.

Today’s corporate education reformers express frustration with the continuity of traditional schooling methods — though most do not recognize the history to which they are intimately tied, since technological innovation is imagined to be as ahistorical as it is apolitical. In a 2013 Google+ hangout, US Secretary of Education Arne Duncan told Sal Khan:

We have to continue to accelerate. The fact that we’re still teaching with a nineteenth-century model makes no sense whatsoever, with twenty-five or thirty kids sitting in rows learning the same thing at the same time, same pace. It’s like Neanderthal. It makes no sense. This idea with technology being a great thing to empower moving from seat time to competency — I don’t want to know how long you sat there, I want to know, do you know the materials? Do you know Algebra or Biology or Chemistry or Physics? If you know it, you shouldn’t have to sit there.

Edward Thorndike, the behavioral psychologist known for introducing scientific methods into the field of education, shared this frustration when he first theorized the possibility of a teaching machine. Textbooks, he observed in 1912, prod a student towards reasoning, but are unable to manage the process of elucidating just enough to help a student arrive at his or her own conclusions.

Described by colleagues as prodigious, efficient, and an extremely rapid reader who liked to read books in one sitting, smoking cigarettes between chapters, Thorndike was preoccupied throughout his career with the quantification of human intelligence — he would go on to create an aptitude test used by the American military during World War I, as well as college entrance exams — but his objection to the use of textbooks in classrooms is an argument against standardization, or at least, against learning at a single standard pace mediated by a teacher.

Thorndike envisioned a future in which texts were capable of offering a self-directed learning experience for schoolchildren: if, “by a miracle of mechanical ingenuity,” he wrote, a book could be arranged to hide information and display it step-by-step, so that page two was only accessible upon mastery of page one, “much that now requires personal instruction could be managed by print” — effectively making the teacher-as-guide obsolete.

Four decades later, B. F. Skinner, a man who neither believed in free will nor had hope for the world’s salvation, stood in front of a new kind of classroom and announced that the future was here. Skinner had been influenced by the work of Sidney Pressey, a psychologist who, following Thorndike’s research on the retention of information through practice, developed a machine he believed would generate an industrial revolution in education (Pressey himself was deterred by the Great Depression).

“I am B. F. Skinner, Professor of Psychology at Harvard University. I should like to discuss some of the reasons why studying with a teaching machine is often dramatically effective,” he announces in a video from 1954.

On the screen, we see an enormous group of teenage children sitting elbow-to-elbow at long tables, rapidly and silently inputting answers into a device that looks like a cross between a typewriter and a record player. In the window of each child’s machine is an incomplete sentence or an equation missing a piece. Once the student fills in the blanks, the machine confirms or corrects the answer. Every child works alone.

“The machine you have just seen in use . . . is a great improvement over the system in which papers are corrected by a teacher where the student must wait perhaps until another day to learn whether or not what he has written is right. Such immediate knowledge . . . most rapidly [leads to] the formation of correct behavior,” Skinner reflects.

Skinner was not only concerned with increasing the efficiency of knowledge absorption for the individual learner, but also for the group. He leaves us with this: “With techniques in which a whole class is forced to move forward together, the bright student wastes time waiting for the others to catch up, and the slow student, who may not be inferior in any other respect, is forced to go too fast. . . . A student who is learning by machine moves at the rate that is most effective for him.”

For Skinner, as well as for corporate education reformers, knowledge is static and students are passive recipients; efficient transmission of information is the goal of education. And technology is the means by which we make the transmission process faster, cheaper, smarter. Gifted children are best served by moving individually at their own pace, “slow students” move at theirs, all in isolation.

This way of conceptualizing learning corresponds neatly with our present economic system, in which individuals either stand or fall on their merits, but it fails to deal with — in fact it conceals — the contentiousness of reality.

Skinner’s new classroom went through many iterations over the decades that followed. A more sophisticated version known as Individually Prescribed Instruction (IPI) was used by students at Pittsburgh’s Oakleaf Elementary in 1965, and described by a contemporary journal of education as “the nation’s first successful operation of individualized instruction on a systematic, step-by-step basis.” His teaching machine, however, was never adopted on a mass scale in American public schools.

Part of the resistance to the technology came from educators. Newly professionalized, they were adamantly opposed to having their role transformed into that of a coordinator. Rodney Tillman, Dean of the George Washington University School of Education, wrote in an essay titled simply “Do Schools Need IPI? No!” that the functions of a teacher using the system are limited to “writing prescriptions for courses of study, diagnosing student difficulties, and tutoring. . . . These I cannot accept.”

Tillman was not resistant to the use of technology in schools so much as he was hostile to the particular vision of learning implicit in teaching machines, which rewarded rote mastery while evaluating student performance in isolation. The skills required to prepare children for the future were, he argued, not didactic, but interpersonal.

And even in neurotic post-Sputnik America, parents tended to share a belief in the broadly humanist model of education. In 1960, the National Education Association (NEA) found it necessary to release a statement reassuring concerned mothers that while mechanical aids were now part of a modern classroom, they would never be the mode of instruction. “NEA Allays Parent Fears on Robot Teacher” was the headline in the Oakland Tribune.

Anxiety about technology in classrooms, or about robots raising the children, was crystallized in pop culture. The Jetsons, which premiered in 1962, is the story of a typical nuclear family in the year 2062. George Jetson works a few hours a week at Spacely’s Space Sprockets, Jane Jetson is a homemaker, and young Elroy Jetson’s teacher is a robot named Ms Brainmocker.

By 1981, at the end of his life, Skinner recanted his belief that technology could solve the world’s problems, observing bitterly that no one had had the inclination to use the tools he’d created. Skinner was not alone in his desire to radically transform education for a new century, or in his eventual disillusionment with this project. Just a few decades prior to the development of the teaching machine, Thomas Edison had declared that books were obsolete and motion pictures would initiate a revolution of the school system within ten years — a process that is still dramatically incomplete over a hundred years later.

The possibilities of education technology remain ambiguous. The tools with which we learn are neither intrinsically empowering, as Skinner assumed and Arne Duncan continues to assume, or inherently threatening. They can be used in ways that are liberating or oppressive. But the popular idea that technological innovation is cruel (Ms Brainmocker) is not irrational.

“Innovation” is almost always invoked by elites to ignore class conflict, to the point that some leftists have come to be wrongly but understandably suspicious of modernization altogether. Experts from Edison onward called enthusiastically for the incorporation of film and radio in classrooms without accounting for the fact that, as historian David Tyack points out, there were still tens of thousands of American schools that lacked electricity well into the 1960s. Of course, these schools were not evenly distributed across the country. They were the ones attended by working-class children, particularly in communities of color.

The Optimism of Billionaires

In 1966, an MIT professor lamented that it had been easier to put a man on the moon than to reform public schools. Today, SpaceX CEO Elon Musk wants to replace the US space shuttle program and blow up education by turning it into a game and adding special effects.

“Give kids a chance to fly,” Duncan said to Khan in their Google hangout. “Let them find their passion and they’ll go to the moon with that.” Why are two such disparate concepts as education and space travel so intricately linked in our public discourse? Education and space are both metonyms for the future.

When today’s children grow into tomorrow’s adults, holding meetings in holodecks and beaming themselves through the galaxy in maroon turtlenecks, they will have replaced us. When science fiction becomes reality, we will all be dead, unless we figure out a way to bring about the impossible.

From the perspective of the tech industry, education and space travel are alike because they are problems in search of rational, personalized, twenty-first century answers, like those arrived at by design thinking. The expectation is that these answers will obliterate material limitations, class struggle — history, past and present.

Design thinking, embraced by key figures in business and especially in the tech industry, insists that educators adopt a perpetually optimistic attitude because that is what it takes to believe everything will turn out okay if we just work together to streamline our efforts. That is what it takes to believe that the best idea is the one that survives group discussion and is adopted. The rabid optimism of the techno-utopian vernacular, with its metaphors that no longer register as metaphors, obscures the market imperatives behind the industry’s vision for the future.

This is intentional. Conflating the future with unambiguous, universal progress puts us all on equal footing. Participating as a citizen in this framework consists of donating your dollar, tweeting your support, wearing your wristband, vowing not to be complacent.

Critiquing the solution only impedes the eventual discovery of the solution. And why make demands for power if you yourself are empowered? Empowerment, as Duncan uses it, is a euphemism. Anger is empowering, frustration is empowering, critique is empowering. Competence is not empowering.

The fact is, education is not a design problem with a technical solution. It is nothing like building a spaceship. It is a social and political project that the neoliberal imagination insists on innovating out of existence. The most significant challenges faced today in education are not natural obstacles to be overcome by increasing productivity — they are man-made struggles over how resources are allocated.

In a frequently cited policy report on academic performance and spending over the past forty years, Andrew J. Coulson of the Cato Institute concludes that dramatic increases in education funding have not resulted in improvements in student performance.

“In virtually every other field,” Coulson notes, “productivity has risen over this period thanks to the adoption of countless technological advances — advances that, in many cases, would seem ideally suited to facilitating learning. And yet, surrounded by this torrent of progress, education has remained anchored to the riverbed, watching the rest of the world push past it.” What Coulson and others who repeat this myth ignore is who specifically is left out of the tech world’s ecstatic march towards progress, and how and why they are left out.

The United States is one of few OECD countries where schools that serve affluent families have more resources than schools that serve poor families. A 2010 OECD report noted:

In 16 OECD countries, more teachers are allocated to disadvantaged schools to reduce the student-teacher ratio, with the objective of moderating disadvantage (OECD, 2010). This is particularly the case in Belgium, Italy, Ireland, Spain, Estonia, Iceland, Portugal, Japan, the Netherlands and Korea. Only in Turkey, Slovenia, Israel and the United States are disadvantaged schools characterised by a higher student-teacher ratio.

In 2013, Andreas Schleicher, who runs the OECD’s international educational assessments, told the New York Times: “The bottom line is that the vast majority of OECD countries either invest equally into every student or disproportionately more into disadvantaged students. The US is one of the few countries doing the opposite.”

In a country where the top 20 percent of the population earns eight times as much as the bottom 20 percent, this inevitably leads to two distinct and parallel systems of education, one for the rich and one for the poor. It’s not that “money doesn’t matter” for reforming the education system, or that technology can be a substitute, but that children from working-class and poor families score lower on standardized test scores than their wealthy peers — and America has many more poor families than rich.

The Cynicism of Managers

Sal Khan’s Khan Academy, funded by generous grants from the Gates Foundation, is the miracle of mechanical ingenuity that Thorndike dreamed of a century ago.

When I first logged on to Khan Academy, I was surprised to find that despite all the tech-industry backing, it is not attractive, simple, or intuitive. Users mouse over the Subjects bar and choose Math, Science, Economics and Finance, Arts and Humanities, Computing, Test Prep, or Partner Content. Clicking on a Math “mission” brings you to a page of basic exercises. In instructional videos, Khan is awkward, a one-time mathlete with a slight twang and the affected exuberance of someone who has been teased but ultimately rewarded for being himself.

The website is interactive in the most mechanistic sense of the word: it provides individual feedback. After ten correct answers, the user can move on to the next concept. Ten correct answers is applied uniformly throughout the site as a metric, though it’s unclear why success in this metric indicates mastery, just as the 85 percent correct required by the IPI system seemed to be arbitrarily selected in order to enable the teaching machine to function. Badges, which are meant to be incentives, are exactly the kind of thing an “unabashedly geeky” adult would think a kid might find interesting.

It is a cloud-based, portable version of Skinner’s teaching machine. Its strength is that it is self-guided: exercises allow repetition and provide students with immediate feedback as they practice.

Memory performance improves with practice, and practice leads to automaticity, which frees up working memory and allows us to concentrate on comprehension. That’s why it’s impossible to gain complex insight into the abstract concepts of literature or algebra until we can read words and equations fluently. Passive practice does not actually improve our recall of information, and Thorndike, who saw the mind as a group of habits, was the first to identify the use of feedback as essential to successful learning.

But where’s the revolution? Khan is quick to say his videos are not a replacement for teachers, a claim that seems disingenuous given that the mission of his project is to “provide a free world-class education for anyone, anywhere.” Pedagogically, the videos are unambitious. Even with a paper textbook, a student can move at his or her own pace and receive feedback by checking answers at the back of the book. Why should a digitized version create a significantly different outcome?

Khan Academy is a fine way to practice math problems or learn a didactic skill. What it is not is innovative in pedagogy or design. As a system of education it is a failure. It degrades both student and teacher by deemphasizing the importance of interpretation and critique in education, just like design thinking does.

One example of the importance of this kind of flexible and evolving practice — especially for children from low-income families — comes from Lisa Delpit, educator and author of Other People’s Children. In talks, Delpit uses a situation she witnessed in a preschool in which a teacher handed out a tray of candy and instructed children to each take a piece and pass on the tray. Some of the children took multiple pieces, and there was not enough to go around.

A teacher evaluating the children without interpreting the context, like a machine, would conclude that the children did not successfully complete the task and need more practice in sharing. In fact, after asking why the children took extra pieces, the human teacher found that they were simply engaging in a different kind of creative economy, saving up a couple of pieces to take home to siblings later.

I suspect the innovation Gates is investing in is not a technological one, but a managerial one. The only truly novel thing Sal Khan has done is produce a cheap and popular way to distribute basic lectures and exercises to a large number of people who like them.

It’s possible that what Gates admires most about him is that one man can teach so many different subjects at different levels, from kindergarten math to cell biology to financial markets. At the Aspen Ideas festival, Gates praised Khan for moving “about 160 IQ points from the hedge-fund category to the teaching-many-people-in-a-leveraged-way category.” Look, he seems to be saying, at all the value that can be extracted from one employee!

In a November 2012 interview, Gates told Fareed Zakaria, “When you revolutionize education, you’re taking the very mechanism of how people become smarter and do new things and you’re priming the pump for so many incredible things. Over the next decade at all levels in all countries, that’s going to change quite dramatically.” Technology “will take that space at the current investment levels and allow us to do a far better job.”

Elsewhere, Gates has called for austerity in public education, repeating the familiar argument that for thirty years we’ve been spending money while performance by American children remains flat. What we need to do, he says, is raise performance without spending more by changing the way money is spent. To that effect, Arne Duncan asked a room full of Silicon Valley entrepreneurs and investors last year, “Can we find ways to scale the amazing teachers we do have?” Systems that “scale” retain quality under an increased workload. Modifying teachers to scale would mean replacing them with robots or computers.

Managers are incentivized to outsource redundant jobs and tasks, but in the past thirty years there’s been a special focus on chipping away at the security and esteem of teachers and the American school system. Certainly it’s about money, as it always is, but the financial backing of the Gates Foundation is astronomical enough that the question is less about actual scarcity and more about how the funds will be spent.

The firing and disciplining of teachers is also an ideological choice: teachers threaten the ruling class. Though they are atomized as workers into separate classrooms and competing districts, teachers are, as Beverly Silver puts it, strategically located in the social division of labor. If they don’t go to work, no one can — or at least, no one with children to look after. As caretakers, teachers are by definition important and trusted community figures, public care workers who can shut down private production.

In the United States, where the vast majority of families continue to rate their own child’s teacher highly, even while believing the political mantra that the nation’s education system is rapidly deteriorating — unique job protections like tenure serve to further strengthen teachers’ capacity to resist neoliberal reforms.

In the same vein, schools are public spaces in which children and teenagers can put down their pencils or laptops or iPads and organize against state violence and coercion, as we’ve seen in the aftermath of Michael Brown’s murder. The possibilities for confronting injustice are so powerful that children (especially black and brown children, but increasingly, all American children) are literally policed and considered suspects in their own school buildings.

Teachers who encourage resistance are essential sources of support and guidance for kids. People do not learn to think critically and construct meaning in isolation — which is the assumption behind the trend of textbooks that respond individually to each student and allow them to move at their own pace. People argue, discuss, play, experiment, and converse. And, as Delpit writes:

Only those who are authentically and critically literate can become the independently thinking citizens required for any society’s evolution. The opportunity to achieve such levels of literacy is even more critical for those whom the larger society stigmatizes. . . . When people of color are taught to accept uncritically texts and histories that reinforce their marginalized position in society, they easily learn never to question their position.

Learning as a group is not a painless process. A good teacher knows her students well, respects them and earns their respect in return, and challenges them to aim for the highest reaches of what Vygotsky called “the zone of proximal development” — their potential.

As Katherine McKittrick has pointed out in response to the idea of trigger warnings being placed on college syllabi: the classroom isn’t safe. It should not be safe. Teaching, for McKittrick, is a “day-to-day skirmish,” and teachers must work hard to create classroom conversations “that work out how knowledge is linked to an ongoing struggle to end violence,” to engage with the history that students bring with them into the classroom and resist reification of oppressive thinking in practical ways.

This winter, during the Hour of Code sponsored by the tech industry and supported by the US Department of Education, Susan DuFresne, a kindergarten teacher and former teacher’s aid with forty years of experience told me, “Children are not standard. They need unstructured play indoors and out to develop skills” like sharing, listening, cooperation, and self-regulation.

The Hour of Code is a publicity stunt in which public school children from preschool up are given laptops and taught to code. DuFresne was vocally opposed. Kids “have different learning styles,” she said. “Some learn faster with technology. But now children as early as third grade will be required to type written answers into text boxes, click and drag, and use multiple tech software tools on the Common Core tests.” Still, her resistance had little to do with fear of new tools, and everything to do with the conceptualization of the role of technology in the classroom.

Another high school teacher, Brooke Carey, who has been working for over a decade in the New York City school system, agreed that technology is often used in public-school classrooms in “a fairly traditional way,” with iPads serving as a fancier version of pen and paper and Smartboards functioning as computerized chalkboards or dry-erase boards. In American public schools, teaching tools have been digitized and optimized for efficiency, but the content and philosophy remain the same.

Even Google engineers know this. An article in the New York Times reported on the popularity of the Waldorf model of education in Silicon Valley as if it were a contradiction: “A Silicon Valley School That Doesn’t Compute.”

Waldorf schools incorporate creative and tactile experiences and tools including hammers and nails, knives, knitting needles, and mud — but not computers — into the curriculum. Engagement comes from the connection between children and their teachers, who stress critical thinking and aim to create interesting, inquiry-based lesson plans.

According to the Times, employees at Google, Apple, Yahoo, Hewlett-Packard and eBay send their children to the Waldorf School of the Peninsula. “The idea that an app on an iPad can better teach my kids to read or do arithmetic, that’s ridiculous,” Alan Eagle, a Google communications executive who’s written speeches for Eric Schmidt, told the Times.

The great irony is that the very Silicon Valley reformers promoting and funding techno-utopian models for American schoolchildren refuse to submit their own children to anything like it, choosing innovative pedagogical models instead of newer touch screens.

The Classroom of the Future

One of the most powerful moments for me as a beginner teacher was seeing a video of a lesson I gave. The recording enabled me to transcend biology: to get out of my own head and see myself as my students did, to notice and interpret rustlings and undercurrents that would have otherwise escaped me entirely due to purely physical limitations.

In an hour, I learned more about my practice than I had during months of supervisor evaluations. iPads are more than glorified expensive dry-erase boards. They could be used to connect teachers, who traditionally operate within the confines of their own individual classrooms, to one another for professional development and growth purposes. Why not film the lessons of experienced teachers and compile a national or global library of what an engaging lesson looks like, immediately accessible to new teachers?

What the current conversation about designing the classrooms of the twenty-first century misses is that innovations do not take place outside of the political economy; they are part of it. What we call technology and what we create with it is determined by the social and political landscape in which it is created. As Marcuse wrote in One-Dimensional Man, “There is no such thing as a purely rational scientific order. The process of technological rationality is a political process.”

For the elite business class, the animating purpose of technology in classrooms is to more efficiently develop human capital, to make some people smarter and faster, and sort out the rest into the discard pile of American capitalism: low-wage labor. Because industrial capitalism makes us all, workers and capitalists alike, dependent on the market for acquisition of the basic necessities of life, we live lives dominated by market imperatives.

The American education system is shaped by those market imperatives — at least for children in public schools. The rich know that JavaScript can be learned in a matter of months. Education for empowerment requires the time-consuming cultivation of a complex understanding of history and one’s place in it, as well as how it continues to shape our relationships and political economy.

When we imagine successful teaching as instruction of X number of people achieving Y level of fluency, we redefine it — whether done by human or machine — from a social (and potentially political) to a merely technical act.

Teachers must continue to be able to help children think critically about the ways that reality is reshaped by technology and changes in the mode of production. How will children who take Google for granted understand research and inquiry? What will friendship be like for children of the electronic age, who have the option of never losing contact with childhood friends thanks to Facebook? Who wins and who loses by the adoption of specific technologies?

It’s impossible to say today how we should teach and learn about social relations mediated by technology, since that is something that must be shaped by praxis — teachers and students working together. But just to imagine the evolution of education in this way is to ask radical questions, beginning with the forbidden one, “What’s wrong with education today?” That question inevitably leads to an even bigger and more dangerous one — what’s wrong with society?

In 1922, a journalist described the way technology changes our relationship to the world: “To the schoolboy of the year 1995 history will not merely be something to be memorized out of books. It will be visualized and made real for him by the moving pictures that are being made now. The people of our time will not be mere history book ghosts to this boy but living creatures who smile at him and walk and play and love and hate and work and eat.”

But this isn’t the way we see history in 2015. Today, we see history as a dying field, in a separate sphere from STEM education; its practitioners likened to the last speakers of a lost language bent on preserving it, and devalued in the same way as women’s work: not well-paid. Humanism is regarded as inherently opposed to machines. And yet, as the journalist of 1922 suggested, technology offers us the ability to form connections and experience intimacy with more people, other people, dead and alive, across time and space.

In a contemporary novel about Victorian England, Sarah Waters has her protagonist notice that the most interesting thing about radio as an invention is not the initial shock of hearing voices across space. “It was even more uncanny to take the ear-phones off and realize that the whisper was still going on — to think that it would go on, as passionate as ever, whether one listened in to it or not.”

Over time, technology has transformed the way we relate to each other and the epistemological foundations of society — the way we perceive reality collectively. This is a truly radical opening for socialists, inside the classroom and outside of it. What will we do with it?