At What Cost?

I’m a couple of weeks behind on sharing links, so apologies for the length of this email and for the outdatedness of some of the news therein. But I devoted much of last week, as I noted in Saturday’s missive, to reading the Pope’s encyclical letter, and as such, didn’t manage to read/write/think about much else.

The Pope wasn’t the only person making a major statement about “AI” last week, of course, or even about “AI” and education. Collectively we just cannot shut up about the topic, can we?

On Wednesday, AFT President Randi Weingarten gave a speech at the National Press Club, where she unveiled a ten-point plan “to boost teaching and learning in the AI era.” Among her proposals – many of which were fairly typical union demands, such as focus on students’ well-being, meet their families’ basic needs, adequately fund public education – were several items that demonstrate the union’s recognition of the growing backlash against not just “AI” but ed-tech in general:

1. No screens (including online assessments) for students in prekindergarten through second grade, unless there is a compelling reason, such as to most effectively support a student with special needs.
2. No student-facing AI in elementary schools — not only to prevent harm, but to build children’s skills like relationship-building and persistence. All other student-facing AI, including digital literacy efforts, must be supervised by educators. And until at least age 16, there should be a total ban on so-called “social companion” chatbots, computer programs that simulate human relationships.

Weingarten also called for better safety and privacy in ed-tech products; for the protection of intellectual property and academic integrity; and for research on the effectiveness of “AI” that is not funded by the industry -- the latter more than a little ironic considering the AFT announced last year that, thanks to that very industry’s financial support, it was launching a “National Academy for AI Instruction,” a $23 million effort, funded by OpenAI, Microsoft, and Anthropic to train teachers on how to incorporate “AI” in their work.

The academy’s professional development sessions now underway, and a handful of journalists have attended and reported back on what teachers are being taught. The gist seems to be, no surprise, to “think critically” when using “AI” in the classroom, a commandment which might sound nice but actually means very little. (Verifying that the generative “AI” hasn’t given you incorrect or biased information isn’t really “thinking critically,” for starters; it’s fact-checking.) This is the industry’s version of “‘AI’ literacy,” of course: “think critically” about how you use technology, but in the end, you simply must use it somehow.

One of the problem with the Weingarten’s call for “devices down, eyes up, hands on” (other than being kind of “cringe” – do we still say that?) is this bifurcated approach: “AI” is bad for kids, but good for teachers – as long as they “think critically,” of course. It’s a false divide, I’d argue, as you cannot so neatly carve ed-tech into “tools for teaching” and “tools for learning.” Nor can you separate either of these from the effects of administrative technologies. I can’t believe I’d have to remind the union of this, considering how often it will shout that “teachers’ working conditions are students’ learning conditions.” The ways in which ed-tech has reshaped the classroom, has changed values and practices and meaning, isn’t just because students spend so much time sitting in front of screens, clicking on things. It's because teachers (and principals) do too.

In a recent essay, Andrew Cantarutti tries to outline a clear distinction between technologies for teachers and those for students, and then argue there are different standards for the evaluation of these categories. Teacher-facing tools, he says, should be judged based on their “professional utility” -- do they make the teacher’s job easier, in a nutshell -- while student-facing ones should be evaluated based on their “developmental costs” -- that is, do they aid student learning or do they actually undermine growth or limit in some way a capacity that a student might still be working on.

One can see this sort of distinction in the AFT’s approach to “AI” -- in Weingarten’s 10-point plan versus its teacher academy: “AI” has too high a developmental cost for young children; they should not be exposed to it while teachers work to “build children’s skills like relationship-building and persistence.” But we needn’t worry about these developmental issues among adults, we’re told. When “AI” promises to make a variety of tasks easier, faster, more efficient for teachers and parents, that’s great – “labor-saving,” even. (Except that it's not. It's just more data-work.)

But as Cantarutti builds his case for evaluating teaching and learning tools separately, he soon admits that this framework may be too simplistic: indeed, most ed-tech surely falls somewhere in the middle; most notably, he observes, the now ubiquitous learning management system, which is based on the promise of efficiency and convenience for everyone -- teachers, students, parents, administrators. That is, everyone’s experience of education is shaped by this product; everyone’s behavior, everyone’s expectations, everyone’s communications, everyone’s assessments, everyone’s thinking about thinking, are bent to fit the programmatic constraints of the software.

Most learning management systems are the clearest examples of a tool that fails the clarifying question — offering teacher efficiency by routing student behaviour through the platform. There are fewer lost assignments, streamlined submissions, and the LMS centralizes communication, but students are required to check a feed, receive automated notifications, and submit work through the platform rather than developing their own systems for managing these things. The teacher’s convenience and the student’s developmental cost are not separable under these conditions. They are produced by the same features.

The learning management system has become the platform through which much of the other software that students and teachers (and even parents) use too, and in the case of Google Classroom, is inextricable from the larger “productivity suite” of tools, which as I’ve argued before, have shaped how all of us -- in and out of school -- think about our cognitive processes as products. One might see the “hidden curriculum” here, the ways in which data and output and efficiency and speed and personalization, for example -- all core values of Silicon Valley -- become the goals of schools as well.

The concern about the “developmental cost” for students when it comes to “AI” is hardly misplaced. Study after study has shown that “AI” damages cognition, harms users’ critical, emotional, and ethical capacities, and creates powerful dependencies (the industry apparently doesn’t even mind using the word “addiction” to describe their intentions here). But it’s wildly naive to assume that those same harms will not befell teachers.

“What Do I Need To Get Done That I Don't Have To Think About?” asks historian Timothy Burke, pondering about the sorts of “mindless tasks” he’s supposed to gleefully hand over to “AI.” “This rhetoric drives me nuts because it is frequently offered without concrete existing examples,” he writes. “It’s always a vague, futureward offer made with no evident knowledge about what it is that most people actually do in work or in everyday life. As if, perhaps, the pitch is coming from billionaires who don’t have to do anything tedious except perhaps to order all those kinds of tasks to be done.”

It is mind-boggling to me that anyone, but especially the teachers’ labor union, would argue that any work an educator does is “mindless” or menial, that any work an educator does is the kind of task that one should automate if they don’t want to have to think about it. I’m not saying that teachers aren’t overworked -- good grief. Rather, I want to remind people that software is not a substitute for the kind of structural change necessary to improve everyone’s lives, in and around the classroom.

The kinds of tasks that I hear teachers being encouraged to offload to “AI” -- grading, lesson planning, communication with students and parents, design of handouts and other classroom material, IEPs -- are actually constitutive of the very work. These tasks -- and yes, some of them can be burdensome, time-consuming, annoying as hell -- are how you come to know the content, the community, the classroom, yourself and others. Nothing about teaching and learning should be thoughtless or careless the way in which “AI” promises thoughtlessness and carelessness as-a-service. Education isn’t comprised of tasks that should be automated; this isn’t work that needs to be made faster and cheaper. Teaching and learning are not something to be optimized or engineered like machinery, turned into the very “factory model of education” that Silicon Valley has spent decades inventing and positioning against.

If we’re worried about what the push-button classroom will do to students, we should probably stop demanding teachers become button-pushers as well.

Everyone Needs Better Labor Unions

“Meta will reportedly let employees take 30-minute breaks from its tracking program” via Engadget. You can sort of imagine that this is how many schools are going to handle mandates for "screen-free" time.

Parenting and/as Technology

“Parents’ Consent at the Heart of Ed Tech Lawsuits” via The 74. It’s fascinating to see where and when education reformers invoke “parents’ choice” as the lever for or enemy of change.

"We Did Our Best!" – Meghan O’Gieblyn on the troubling history of parenting metaphors in “AI," and how, with Anthropic, that’s changing – and yikes, perhaps even more troubling!

“Does Your Kid Have iPad Rages?” asks Liz Krieger in The Cut.

In the new Toy Story movie, the child protagonist Bonnie is apparently too captivated by her knock-off iPad to play with Buzz and Woody. I’ll withhold judgement on the movie’s message about “the digital” until I see it. (Who am I kidding. I won’t see it.)

Elsewhere in concerns about boys’ toys, The Telegraph worries about “The terrifying rise of schoolboys making AI girlfriends.” And Education Next reports that “Teen Boys Are Gambling. A Lot.”

Terrible Investments

“AI Doesn’t Have ROI,” Ed Zitron argues ever-so-succinctly. I’m kidding with that adverb. But I’m serious when I wonder what exactly schools/teachers unions think the “ROI” is for them in partnering with this monstrosity?

“Will AI Help Revive the ‘Stale’ OPM Market?” asks Inside Higher Ed.

“A University System Went All In on A.I. Now It’s Tearing Itself Apart,” The New York Times reports.

“The EdTech Backlash Is Here, and It's Just Getting Started” Nora De La Cour argues in Jacobin.

The Brookings Institution’s Rebecca Winthrop has an op-ed in The New York Times on “What 370,000 College Essays Tell Us About A.I.’s Effects on Creativity.” But I didn’t make it past the first sentence, sorry: “I am a big fan of technology.” She lost me...

I also did not read the latest from Todd Feathers in Wired on Alpha School, but if you have the stomach for it, here's the link. I'm sure this time, it's gonna be grrreeeat.

Florida, Man

Via The Wall Street Journal: “OpenAI Sued by Florida’s Attorney General Over AI Harms.”

Elsewhere...

Ben Williamson continues to keep an eye on the other forms of creeping / creepy scientism in ed-tech: “Neuroscience, technology, and translation in education policy and practice.”

Essayer Means “to Try”

"No, Artificial Intelligence is Not Conscious" by Ted Chiang.

“Authenticate thyself” -- Marion Fourcade and Kieran Healy in Aeon:

our sense of who we are is assembled in a strange and tangled fashion. The machinery of ordinalisation attends carefully to individuals rather than coarse classes or groups. By doing so, it appears to liberate people from the constraints of social affiliations and to judge them for their distinctive qualities and contributions. It promises incorporation for the excluded, recognition for the creative, and just rewards for the entrepreneurial. And yet this emancipatory promise is delivered through systems that classify, sort and, above all, rank people with ever-greater precision and on a previously unimaginable scale. The resulting social order is a sort of paradox, characterised by constant tensions between personal freedom and social control, between the subjective elan of inner authenticity and the objective forces of external authentication. It gives rise to a certain way of being, a new kind of self, whose experiences are defined by the push for personal autonomy and the pull of platform dependency.

“Do Not Resign From Life,” LM Sacasas reminds us.

Today’s bird is the red junglefowl, a bird in the phasianid family (which includes pheasants, quail, turkeys, and peafowl). While the bird was first formally described by Carl Linnaeus, bless his heart, it has a very very long history with humans: the red junglefowl is the species that gave rise to what we now think of as “the chicken,” thanks to domestication some 8000 years ago. Unlike chickens, red junglefowl remain wary of people. (No surprise, as they're apparently quite tasty.) According to Wikipedia, the male junglefowl crow “cock-a-doodle-doo” (if they’re crowing in English, at least, which as this species is found throughout Asia, they probably are not).

Thanks for reading Second Breakfast. Please consider becoming a paid subscriber as your financial support enables me to do this work – the hard and important work of eschewing all Tom Hanks movies.