Review by Lee Skallerup Bessette

How you read Audrey Watters’ new book Teaching Machines: The History of Personalized Learning (MIT Press, 2021) will greatly depend on how to you view technology generally and ed-tech more specifically: do you see tech as the great (eventual) savior for our problems, or do you view tech more skeptically? 

Watters, a long-time ed-tech critic, falls firmly on the side of the skeptic. Her blog, Hack Education, is a more than decade-long project that constantly and consistently looked upon ed-tech solutionism and dared to point out that the emperor had no clothes. This has long made her a target of who can only be described as “tech-bros” online, unleashing their ire and anger at her audacity and truth-telling.

Teaching Machines front cover. MIT Press, 2021

Teaching Machines is Watters’ attempt to historize the latest ed-tech buzzword, “personalized learning.” Long before MOOCs, Khan Academy, predictive analytics, and automated interventions, there were the teaching machines. Watters traces the idea, often associated with the Behaviorist B.F. Skinner and the post-WWII mechanical boom, as far back as the 1920s with Sidney Pressey and his concept for the mechanical teacher. That we are still having debates about the possibility of automating teaching and learning 100 years later shows the need for Watters’ important book. 

Watters has written a highly readable yet exhaustively researched history of the failed attempts to successfully invent, produce, and commercialize a teaching machine. These men (and they were almost all men, white men at that) promised that students would finally be able to learn at their own pace and teachers would be free from the drudgery of grading. Learning would finally be an efficient process, improved by the promise of the latest technological advances. Never mind that not one of these self-proclaimed ed-tech saviors had any experience teaching primary or secondary students, or any knowledge of pedagogy, curriculum, or educational politics. 

The striking naivité of these men was particularly galling: the fact that companies weren’t lining up to produce the machines and school districts weren’t clamoring to purchase the machines was baffling to them, given what perceived promise the teaching machines held, promises that the inventors were all to happy to make to anyone who would listen. We are fortunate, in a way, that these men only had access to limited funding from some government bodies and various foundations and not the massive amounts of money venture capitalists now throw around at the next big thing in ed-tech; our classrooms would look much different today. 

I am reminded in reading Watters’ history of the dystopian story, The Evaluation,” by Chad Sansing. This is the future that Skinner and his peers imagine made real, the classroom that I kept picturing while reading the history of the teaching machines. It’s a classroom where the students only work by themselves on machines, and the teacher sticks to an approved script to maximize learning, and the assessments go nowhere except to justify funding. 

What technological solutionism forgets about learning is that it is fundamentally social and that assessment done right isn’t drudgery, but a key in understanding an individual student. There is nothing in any of these men’s plans that would enrich the quality of the human interaction in teaching. It’s like the infamous episode of the satirical show South Park with the underwear gnomes (they steal socks, have a mysterious second step, and then there will be profit!). For those who believe in the power of technology to transform education for the better, that giant question mark is the unanswered question of how the technology will improve the quality of the entire learning environment that includes real, live human beings, filled with the vastness of humanity. 

Substitute “teaching machine” with “app” and just about everything Watters chronicles could have happened yesterday. You could steal the storylines from her history and easily update them for a handful of episodes of the HBO satire, Silicon Valley, without changing much. And while I keep referencing satire to make sense of this history, it is nonetheless a true story, a real history. It reaches the point of parody or satire because itå has been repeated so many times, with the belief that this time, the technology will be good enough, the market will be ready enough, the people and governments want it, and schools and schooling will finally be revolutionized by and through technology. 

You can’t critique her writing, her research, her storytelling, so the only thing left to critique in Watters’ book is her framing of this history, the conclusions she draws. She is not subtle or equivocal in what conclusions she thinks the readers should draw: that we are doomed to keep wasting time, money, and resources on technological solutionism at the expense of our children and their education if we don’t learn from these past attempts. Critics will also point to perceived successes in the current ed-tech marketplace as evidence against her conclusions, but these successes are based on sales and users, not in terms of actual positive educational impact. 

In her conclusion, Watters positions current resistance to ed-tech solutionism with past resistance. She writes:

From the history of refusal, we can see when students and teachers and communities protested attempts to engineer them, they imagined and built – most notably, perhaps Freedom Schools, we can glean ways to construct and share knowledge that depends on humans rather than machines, liberating us from the efficient control of the “Skinner box.” These practices privilege the much messier forms of teaching and learning, forms that are necessarily grounded in freedom and dignity (263).

So if you find yourself asking of Watters’ conclusions the same questions as Skinner and his ilk do during the narrative, then you’ve missed the point of the book entirely, and history of ed-tech and teaching machines will continue to repeat itself.  

Lee Skallerup Bessette is the assistant director for digital learning at the Center for New Designs in Learning and Scholarship (CNDLS) at Georgetown University. You can read more about her work on her website,