Ed-Tech is Management
On the whole, technical people come to share the perspective of those who wield power rather than those over whom the power is wielded, with managers rather than labor, with officers rather than soldiers." – David Noble, Forces of Production
Cybersecurity officials have dubbed it "Black October" – the worst month in computer security history (or the worst so far), as hackers using "AI" have uncovered and exploited vulnerabilities across a range of computer applications and programming languages, including some of the core building blocks of the web. The Department of Homeland Security was hacked. Fan fiction site AO3 was hacked. Education Testing Service was hacked. The Florida Department of Education was hacked. Crypto exchange Binance was hacked. French refrigerators, hacked. The list goes on and on. "AI is changing everything about hacking," writes 404 Media's Jason Koebbler. "No one knows what happens next."
But plenty of people are going to speculate -- to try to shape the narrative, shape the future.
According to ed-tech consultant Phil Hill "what's next" for education will be an acceleration of the movement to "managed services" -- that is the (continued) outsourcing of technological capabilities to third-party providers. Calling this "ed-tech's reckoning," Hill writes that
"I believe we are entering a period that will significantly accelerate the move toward managed services, shared infrastructure, and restrictions on locally operated applications exposed to the public internet, particularly for under-resourced institutions. This reaches into budgets, institutional autonomy, and the ways educational technology gets built and delivered — well beyond the immediate work of a CIO responding to a security alert."
(He then includes an "AI”-generated image of two wrestlers -- Hacker and Ed-tech. The latter, about to get his face smashed with a folding chair labeled "the new AI" appears to be Mark Wahlberg. Um. What?)
I want to build on some of the arguments I made in my talk earlier this week at the launch of Princeton Societal AI and remind folks that there are different ways things can unfold -- different stories, different future for ed-tech. What if educational institutions were to seize this moment (a moment that extends politically and calendrically well beyond "Black October") to strengthen rather than further weaken institutional expertise and capacity and undermine the agency and autonomy of faculty, students, and staff? What if this is an opportunity to ask better questions about why schools (educators, students, staff) build and use technologies rather than just relocate their development and maintenance elsewhere?
Indeed, outsourcing ed-tech to managed services is already what schools, both higher ed and K-12, tend to do -- not just with the student information system or the learning management system, but with the email server and the video conferencing software and the library catalog and the productivity suite and the server hosting and the HR system and the help desk tickets and on and on and on and on. And as the recent breaches of Canvas and PowerSchool (etc etc etc) have demonstrated, this has never been a guarantee of security.
Rather, it is a shift of budgets, of labor, and -- this is key -- of accountability; and perhaps at the end of the day, the latter is what so many managers and executives find appealing about "managed services": the buck stops elsewhere.
There are alternatives, no doubt. But they require a commitment to developing and maintaining human capacity inside of institutions rather than hoping that externalizing the work also externalizes the risk. It means rejecting the notion that "innovation" comes from purchasing a technology product (or as is more likely the case, paying for a subscription) or that "innovators" are found in tech companies rather than right there on campus, in schools.
"Colleges and universities will still make local choices, use open-source software, and build tools for their own needs," Hill admits. "But the ability to put an application on the public internet carries an operating obligation that is becoming much harder to sustain through spare capacity, occasional upgrades, and the knowledge of one helpful colleague." And while I don't want to dismiss or downplay these latest cybersecurity threats -- although to be clear, some of these issues are longstanding and well-known bugs which "AI" didn't "discover" as much as made it easier to brute force these into working -- I do want to remind everyone that "the open Internet is dangerous" has been the selling point for education technology services like the LMS and TurnItIn since the outset. And I don't see how anyone could keep a straight face and argue that these services have made students (or professors or schools) "safer."
There's a part of me (to no one's surprise, I'm sure) that wants to say, sure, shut down the Internet. Let it die. Let's step offline for a few. We could all use little more self-reflection (some of us, a lot more), asking ourselves why we've decided that what's missing in our personal and professional lives is programmatic, why we believe that the hole in our souls can be filled with software -- be it software we build ourselves or buy/borrow/rent from someone else.
And yet I'm incredibly reluctant to endorse any move that makes these decisions for other people, particularly when it coincides with a number of other, ongoing efforts to reduce (or even eliminate) the personal and professional autonomy that has been fundamental to the educational project in the US: the freedom of what to teach and how to teach, the freedom of what to research and how to research.
It's always something when it comes to the efforts to undermine workers' rights: arguments about better efficiency, more productivity, lower costs, whatever. Add to that now, better security. Management always frames workers, no doubt, as a security risk. As such, it will demand more access to workers' computers, their browsers, their downloads, their keystrokes. Historically, control of the machinery has always been a move to deskill labor, with the aim of reducing costs (if not eliminating workers altogether) and gaining complete control of the production process.
That is the goal -- the explicit goal! they're not even trying to hide it! -- of AI companies, particularly when it comes to knowledge work. To outsource the machinery of knowledge work to "managed services" undermines the workers; it undermines the work.
Everything, Elsewhere, All at Once
Google says it will make it easier for schools to restrict students' access to its AI products. But if they opt to do so, Google says it will also turn off their access to Search. LOL. I mean, I think students should learn that there are other search engine products available, some of which do not make money through surveillance and behavioral engineering. And all this on heels of last week's very very bad day, when the company lost almost $500 billion in market value after its quarterly report revealed how "AI" was cannibalizing its ad revenue.
Also hilariously embarrassing: Dartmouth president resigns over his "AI" usage.
"The ‘Human in the Loop’ Is Doing What, Exactly?" by Jeppe Klitgaard Stricker
"The Knowledge and the Knower" by Andrew Cantarutti
"Freedom of Intelligence" by Annie Abrams
"Meta considered a ‘whisper strategy’ to turn academics against Jonathan Haidt," The Verge reports. (danah boyd insists that no one paid her to say that Haidt was wrong. Perhaps not. But it's still worth thinking about how the tech industry has funded very prominent social science and technology researchers. This is always the danger with this sort of big corporate and philanthropic money: funders don't have to explicitly ask you to do or say things.)
Another tragic story from The New York Times on how ChatGPT usage turns deadly: "She Trusted ChatGPT With Her Medication Decisions. It Cost Her Everything."
"Ed Reform is Dead," Ben Riley writes. "It’s Parent Choice or Public Schools."
Let me just remind you of something: just because a pundit (blogger, consultant, influencer, hype-man) says something is happening doesn't mean it's actually a trend. Often he's just wish-casting, hoping that something he likes becomes popular. (Honestly, it's big sad "please clap" energy.) Sometimes he's talked to a couple of buddies who'll agree: oh yeah, that does seem cool. But N = 3 isn't much of a trend either.
I get it. Sometimes if you spend a lot of time online, it can seem like everyone is talking about something that you're not doing. FOMO is a feature, not a bug. It keeps you clicking.
Remember the Cronut? Did you feel left out, back in 2013, because you didn't live near Dominque Ansel's bakery? Were you sad that there wasn't a croissant-donut-combo available in your hometown, that you couldn't stand in line for hours to try one even if you wanted to? Did you still manage to go on with your life, even maybe finding joy in other less glitzy pastry options?
All this to say, what if agents aren't really a thing? (I don't just mean, what if a lot of it's just hype either. What if a lot of it's just dumb: a script, barely even automation let alone autonomous "AI."
It's okay if you're not using one.
The Cronut, on the other hand - that shit is genius.
Today's bird is the magpie, which despite its reputation -- thanks, Rossini -- does not actually like to steal shiny things. These birds are incredibly intelligent -- one of the few non-mammalian species that can pass the mirror test. They also make tools. They can solve puzzles. They work in teams. Indeed, researchers have found that magpies who live in larger social groups are actually smarter than those who live in smaller groups – perhaps more social demands allow for more cognitive demands as well.
Thanks for reading Second Breakfast. Please consider becoming a paid subscriber as your financial support is what allows me to do this work. (And, ya know, eat.)
