Use Google's AI Or Else

Tired of tech companies trying to shove AI down your throat? Or worse, your children’s throats? Google says open wide, or else!

As covered by Dana Goldstein and Natasha Singer for the New York Times, Google has announced that starting next fall, schools will have the option of turning off AI in the Google toolbox offered to students. But--

If the school turns off AI, threatens the not-actual-mobsters that run the company, students will not have access to Google search. As Goldstein and Singer report:

“Search and A.I. are deeply integrated,” said Chris Phillips, vice president for education at Google. If a school has banned A.I., he added, students “will not have access to search, and they’ll see a message from their school that indicates it’s not available.”

And, you know, creating a version of Google for students that doesn't incorporate AI is just incompatible with their plans to recruit young customers too hard to do.

This from the same company that cannot produce a genAI product that doesn't easily facilitate cheating. Guided Learning is supposed to get students to Learn Stuff rather than just using a cheatbot (spoiler alert: it's still just a cheatbot).

But Google, the folks who previously declared internally that its school products were intended to create a pipeline of future users, really really really really REALLY want to put AI in front of the 70 million future users sitting in schools. They have just had a "negative cash flow quarter" thanks to massive AI spending. That is further exacerbated by the discovery that this AI-incorporated search that they want to force on students actually makes less money than the old approach.

Look, if you are of a Certain Age, you might still harbor the quaint notion that Google's main business is running a search engine. It is not; the New Google's (the post "Don't Be Evil" Google) main business is advertising. But we tend to be sensitive about the idea of turning over young minds and eyeballs to advertisers, so it has been useful for Google to wear its Ed Tech mask when approaching schools.

Only the Ed Tech mask won't cover up AI, and educators and parents and even students have been pretty clear that they'd rather not give children unfettered access to AI products. 

This is a hell of a bold move for Google, considering that there are plenty of other search engines in the world.

I expect that many other tech companies will be whipping up a counter-Google education campaigns by end of business today.  I can only hope that this piece of evil thuggery will blow up in Google's piggy corporate faces, and I certainly intend to share with my own local school district how I think they should respond. 

In My County, Children Vote

I will vote today. I always vote; I am what political operatives call a super-voter, a person who has voted in every election he was ever eligible to vote in, and therefore can be counted on to vote next time, so aiming a megaphone of electioneering at him might not be a waste of time. Today, I only vote in three races, and none of them have any great significance because it would be a huge shock and surprise if they didn't go the way they are expected to go. Josh Shapiro seems hard to beat, and my state and US House members-- well, in my area, if Jesus registered as a Democrat and ran against a drowned rat registered as a Republican, I would not bet against the rat.

Anyway, I will still vote, because that's what good citizens do-- they vote. No matter what. No matter the outcome.

But I won't vote until after I pick up the Board of Directors from school, because they always vote with me. Not in the actual election. But in my county, someone decided years ago to start offering ballots to children.

The children vote on issues like favorite type of cookie or best pet or best way to spend vacation. They vote on a special ballot, hand it in, and in a week or so, the local newspaper/website will publish the results. 

One actual result of this is that the Board knows what an election is, how it the voting place is laid out, how one approaches it, waits one's turn, signs your name, marks a ballot-- the whole business. When the day comes that they can vote in the real election, it will be exciting, but it will also be familiar and unintimidating.

Yes, their teachers sometimes run election simulations. But this gives kids exposure to the real thing, the whole business of voting. No mystery. No fear.

And as a bonus, thanks to some of my former colleagues, some high school students are spending the day as volunteers at polling places.

The cost of this Young Voter Education Program is minimal. There's some non-zero number of man-hours used to tabulate. There's the copier paper used to make the ballots. That's about it. 

So my hat's off today to county officials who hit on a simple but effective method of educating the next generation of voters about how the process works and helping them prepare for the day when they can try to execute their obligations as US citizens. 

It's Not A Tutor

There's a new ten-week study from the UK looking at AI as--well, the study people writing about it (like Greg Toppo at The 74) insist on calling it AI tutoring. There's also a recent paper studying the use of Khanmigo which is also characterized as an "AI tutor." We'll look at the results in a moment, but first, I want to look at that term--

"AI Tutor" is yet another marketing term. "Tutor" tries to connect the software to the (somewhat dubious) data showing that tutoring does awesome things for learning! (Two sigma!!) But the use of "tutor" also evokes the idea of a human tutor, a knowledgeable human expert who sits down with the student to provide face-to-face individual support and assistance for learning. That's such a more charming image than, say, a solitary student hunched over a desk and working her way through a ten-inch stack of worksheets.

In fact, if we're thinking about a student ploughing through worksheets or simply buried in a book about the topic at hand, nobody says that students is being tutored. They're just studying. The worksheets are a study aid. The book is just a study aid. And any sort of software-- that's just a study aid being viewed through a computer screen. 

The advantage of actual tutoring (if any) is working with a human who can enter into dialogue about what the student does or does not understand, drilling down to find the mental error the student is tripping over. The tutor also has the advantage of the ability to put human-to-human pressure on the student to perform-- to actually try and keep at it, as opposed to just checking out and/or wandering off. 

GenAI promises that kind of interaction, or at least markets itself on that notion. But the Oreopoulus and Low Khanmigo study found, as did an earlier study this year, that students don't actual interact in that way. Students are supposed to be able to message Khanmigo to discuss issues, but only 17% followed up on a mistake. And many didn't do anything at all. As Khan himself told Matt Barnum of the program's tutoring efficacy for students, “They just didn’t use it much.”

The new study is a trial with ten weeks worth of lessons for 1,179 UK students with Eedi's product, built on Gemini, Google's fine-tuned for teaching genAI bot. Google Deepmind did the analysis and discovered that the AI didn't make much difference compared to "static content." There are differences between girls and boys when it came to AI vs. human tutors, but generally not much difference. Eedi proudly notes that the AI generated hardly any errors. But in terms of advantages, it offered just one-- it's probably cheaper than an actual human tutor. 

What about the 5% problem-- the finding that these programs work great for students who use the program as intended, but that's only about 5% of all students. And there's plenty of reason to think that the 5% students are the ones least in need of help or at the very least would have done just fine with any sort of study aid. Eedi's students engaged more regularly, however there is zero mystery to that finding-- the Eedi sessions were built into regular class time. And 1 in 5 students still ignored the software when a "do you need help" message popped up. Clippy can relate.

AI "tutors" still lack one other element of the human approach-- motivation. Bibi Groot, Eedi's chief impact officer, told Toppo that AI tutoring is “great to support students who are already motivated, but we need to figure out how to motivate students who aren’t motivated in the first place.”

Look, it's not a tutor. It just isn't. It's one more study tool, and for the right student in the right situation with the right material, it might even be a useful one. But a study tool is all it is, and the rush to make it a useful one could only be aided by dropping the charade that it is a replacement for a (costly) human tutor and instead focusing on how it could be a more useful study tool.

But the AI industry isn't going to do that, because it would involve going back to what AI "tutoring" really is-- an extension of the old ed tech Personalized Adaptive Learning model, which was mostly just an endless series of worksheets chosen by an algorithm and fired off at some beleaguered student until they were allowed to unclamp themselves from a screen. But an infinite pile of worksheets was never very compelling-- not even if you shout "But they're coming at you through a computer screen!" 

AI ramps up some of the tech capabilities ("Now, instead of selecting from a library of thousands of worksheets, we can create all new worksheet based on our library of thousands of worksheets!") and the ability to mimic human speech (but not the human thought or insight behind it-- being able to generate a question doesn't make you Socrates). But a study guide, however zippy, is not a tutor. 

ICYMI: Sugar Crash Edition (11/1)

Happy Day After. Now please, don't put up your Christmas decorations today. 

Here comes this week's list. 

Why Gov. Shapiro must say no to federal vouchers

A retired PA superintendent explains why the federal vouchers are a lousy idea.

Public Schools Are Prepping for Federal Choice Cash. They Still Might Not Get It

At EdWeek, Mark Lieberman addresses the continuing pipe dream that the federal vouchers will somehow infuse public schools with cash.

Empathy. Again.

Nancy Flanagan wonders why empathy is in the cross hairs again.

Commissioner Mack says competition is improving EVERYTHING. Show Us The Data.

Sue Kingery Woltanski looks at Florida's education honcho's delight in the privatization of education.

Was American education ever great? What Linda McMahon gets wrong about the past

Matt Barnum provides some historical perspective on Linda McMahon's ahistorical ranting.

It’s Time to Teach Students to Be Drivers of AI, Not Passengers

At The 74, two Newark teachers describe how they manage AI in their classrooms, and it's not the worst framework I've seen.

University of Minnesota report warns student AI use poses 'severe' threat to learning

More research to fill the vast gap that these researchers decry. Some of the conclusions are blindingly obvious (AI use for primary students is a bad idea), but hey, now there's actual research to back it up.

There Is No Alabama Miracle (But There Is a Lesson)

Bruce Baker explains why Alabama's results are nothing to get excited about.

Why Students Hide Themselves from Us

Alfie Kohn, on why students don't always let teachers or parents see who they are. 

One to the Other

Adrian Neibauer shares a learning experience with his students based on "The Monsters Are Due on Maple Street."

Ed Reform is Dead — it’s Parent Choice or Public Schools

Ben Riley and Jame Shanahan offer a take on the history of modern ed reform. I'm not sure I entirely agree with it, but it's an interesting version of the history and though-provoking at that.

Ohio’s Center for Christian Virtue Plans National Network of Religious Private Schools Paid for by Federal Tuition-Tax Credit Vouchers

Jan Resseger looks at one of the groups hoping to cash in on federal vouchers.

Selling Stories

Audrey Watters has a lot this time, including AI marketing, Democratic education leadership, and a host of valuable links.

Who Gets To Tell The Stories About Our Schools?

Jose Luis Vilson looks at the people telling all the stories about schools, and what it takes to have your story listened to.

The Slop is Creeping In

Michelle Banks writes about the frustration of arts communities rejecting AI "art" but still using those execrable AI posters for publicity.

More Than Music

I play in a community band, and a couple of my musical colleagues play in a larger community band festival. The conductor of their last one, Jari Villaueva, wrote this lovely tribute to community bands.

This week at Forbes.com, I reviewed Jen Jennings' excellent new book and also looked at the growing crisis in forced downsizing for districts.

Donnie Iris passed away this week. He was a Western PA icon of sorts, a regional star who peeked around the edges of national fame. This song is the closest he came to a real national hit (though the album King Cool has much of his best work).

The Imaginary One Size Fits All Education

It is one of the most-repeated myths of public education critics. Here's Linda McMahon spouting it again last summer:

We reject the one size fits all education approach. Parents, states, and local leaders know what’s best for our students! The Trump Administration is returning education decisions to those closest to our children. https://t.co/ItvHS7JxvB

— Secretary Linda McMahon (@EDSecMcMahon) July 10, 2026

"We reject the one size fits all education approach." 

You and everyone else, ma'm. The American school system is not one size fits all. It is structured in a way that makes one size fits all functionally impossible. There are roughly 13,000 public school districts (plus private schools, plus charter schools) and each is controlled by its own local board, with each board making its own set of decisions that are then filtered down through individual building administrations. That's a whole lot of different sizes of education.

Which was, of course, the central complaint of the standards movement-- that there were too many districts and buildings and teachers doing all sorts of different things, and why don't we impose one size for all? Common Core was their best, biggest attempt to inflict OSFA on the nation, and it failed, a big clunky wave that shattered and fragmented on the individual rocks of a fragmented system. People with power tried, really hard, to make one size fit all in US education, and it didn't work, not even when the standards went into witness protection and took up residence in many states under an assumed name. 

One size fits all for education in this country??!! When I was still teaching, I wasn't even one size fits alling it for a single day. Each class needed its own fit. My vocational students didn't need to master MLA research format and my college bound students didn't need to wrestle (yet) with resumes and job reports. Certain students needed a different approach, and certain other students needed an even more different approach. And that's before we get to the occasional special class, like the class I once taught that was 75% teen moms. 

Nor was I one size fits all from year to year-- every year involved more fine tuning and editing and reframing what I had done before. 

One size fits all schooling is right up there with hasn't changed in a century schooling and factory training schooling-- they are all glib criticisms effortlessly deployed by people who don't actually know what goes on inside actual schools. These are also the same folks who think that one size fits all schooling is a great idea if they get to pick the size ("If all teachers did X, then all students would learn Y") and then get pissed off that they can't make the same t-shirt fit on every single teacher and student in the country. 

One size fits all education is, in fact, a terrible idea. Fortunately, it's an idea to which education is highly resistant. Unfortunately, it's an idea that amateur education reformsters never tire of A) leveling as an accusation or B) trying to impose. 

AI Not Paying Off Outside Tech Sector

Yeah, maybe not

The chief economist at a major "asset management" company says that AI is not paying off anywhere outside the tech sector. Nobody is benefiting from the damned botslop except the companies behind it.

Torsten Slok is Partner and Chief Economist for Apollo Global Management, an American asset management outfit that handles about $850 billion. In a post earlier this month, Slok makes a point of separating the Magnificent 7 tech companies (NVIDIA, Apple, Microsoft, Alphabet, Amazon, Tesla, and Meta) and the S&P 493 (what you get when you remove the big tech companies from the S&P 500). 

Slok's statement is pretty blunt: 

Despite massive investments and widespread hype around artificial intelligence, the technology has delivered zero financial payoff to corporate America outside of tech.

 No revenue per worker increases, no profit margin increases, and no productivity boosts for the S&P 493-- only the for the Mag 7 (that is apparently what we're calling them in the investment world). 

Can American corporations turn cheap "intelligence" into revenue growth and increased profits? Slok has his doubts:

AI may be a great calculator, but in business, intelligence is mostly judgment. Judgment depends on context, and the answer varies with which human or which model you ask. Perhaps most importantly, AI does not understand accountability, trust, culture, unwritten rules, ethical trade-offs or fairness, all of which shape reputational risk and ultimately drive revenue growth and profit margins.

Emphasis his.

What a very human way to look at AI's utility (or lack thereof). The article comes with a series of charts that make his point with simple graphic clarity. 

Two points for the education world to consider. AI is failing to transform any sector outside of its own, which suggests that its ability to transform education is being grossly overstated. It also suggests that the inevitability argument-- that students must learn to work with AI or be left behind in the AI-centered future-- is being grossly overstated. For the zillionth time, education leaders would be wise to work a little harder at distinguishing between legitimate attempts to predict the future from disingenuous attempts to market a particular future into existence. 

Bad Writing Advice for Day of AI

If you are a high school teacher looking to incorporate genAI into your writing instruction, MIT has some really bad advice for you.

You may recall Hour of Code, the giant corporate-infested campaign to popularize coding (if not, I recommend Natasha Singer's Coding Kids). That has now become the Hour of AI, but apparently some folks want to extend that a bit. Powered by MIT RAISE and the "leading researchers in AI education," here comes the Day of AI, which comes equipped with a whole raft of units more appropriate for actually many several days of AI. Incidentally, RAISE stands for "Responsible AI for Social Empowerment and Education," so there's a sign of where this is headed.

The list of curricula is extensive and runs from pre-K through 12. For instance, there's a K unit "Machines All Around Us" that "introduces students to the foundational idea that machines are human-made tools designed to help people do work and solve problems." There are units about how to verify information and thinking critically about AI, and they are not at all terrible (not perfect, but not terrible). 

However, "The Writing Process in the Age of AI" for grades 9-12 is hugely problematic. 

Let's start with calling this the "Age of AI," as if AI is the dominant feature of our times. I'm willing to entertain an argument about that designation, but I'm not prepared to concede AI world domination as a fait accompli. I'm not able to see the educator guide or student resources (because I'm not going to give them my info), but I can see the slide deck that accompanies the week's worth of lessons. So let's take a look.

Lesson 1 is "Demystifying the Machine," and it does a nice job of making clear that LLMs do not use "thinking," but rather a process of predicting the next word. It sets up LLM prediction against human judgment, and that's fair. It also provides an age-appropriate but reasonably accurate explanation of what LLMs actually do. So we're actually off to a good start. It's not going to last.

Lesson 2 is "from ideas to argument." This section highlights where the first place the creators think AI fits in the writing process, and it begins with the dubious advice that "AI is a good tool for brainstorming, questioning, and organizing." It adds the caveat that AI can give you more ideas but can't decide which ideas matter. The slides suggest that you prompt the AI, then annotate the results, weed out the ones that are weak or with which you disagree. They do not add "get rid of ideas that are actually wrong," which strikes me as the most time-consuming part of this approach. The rewrite the one you pick. 

"Choosing, revising, and rejecting" are "where the thinking is" is the claim here, and while I agree that that's where some think could happen, I have seen this movie before in the classic shake and bake high school "research" paper. Student locates however many sources you required, looks for the one that's easiest to turn into the assignment, and tries to rewrite something so that it says the same thing in a non-plagiarizey way. No thinking is required for any of it (which is why I stopped making that assignment).

Could this AI version of brainstorming involve thinking? Sure. Will it? Probably not.

Lesson 3 is "from prompt to persuasion," and it continues our trip through the writing process, and now we are headed way, way out into the weeds. Ever get stuck getting an essay started? AI can write that draft for you, "especially if you give it a high quality prompt." By "high quality prompt" you should include the sections you want in the essay, the arguments you have developed, counterarguments (if you're doing that, I guess) and other aspects of the essay you want to specify. I am trying to imagine a universe in which a writer has thought through all of the content and organization of an essay and yet somehow the actual writing escapes them. I cannot. If you have developed all the elements of the "high quality prompt," the actual placement of words on the page is barely work at all. 

The creators of this slide show remind you that AI can give you a draft, but you have to revise and rewrite to "make it yours." But let me share this: in 39 years of teaching human high school students, I found that getting them to do meaningful rewrites and revisions on their own rough drafts was by far the hardest part of the writing process. But the creators of this show want students to revise the AI draft "so it is accurate, sounds like you, and says what you want it to say." But if you know what you want it to say, why wouldn't you just start by writing it yourself? 

There's another layer of bad here-- this is training students to be accountability sinks. Instead of being creators themselves, this approach trains them for a role in which they do not create, but accept the responsibility of fixing whatever the bot extruded. 

And of course the reality here is that students will have the AI do everything and make nominal attempts to pretend to "make it their own." This is a framework for formalized cheating.

Lesson 4 is "name the process" which is about documenting the many ways that you used a bot to do the work. The lesson works through the details required-- it's not enough just to say you used AI, but specify which brand bot, record your prompts, and "reflect on how the use of AI affected your writing--this is your chance to show your authorship." You're going to show how you an author by writing insights about the essay that the bot wrote for you. 

Lesson 5 is "deciding when to use AI." This lesson talks about cognitive effort and cognitive debt and encourages students to think about how using AI to avoid thinking is probably bad. It cites research that shows that students who wrote essays with genAI tools could not reproduce a version of those essays later on their own. "It was like their brains hadn't stored the information." No shit. 

The best way for a classroom teacher to use this unit would be to go straight to Lesson 5, maybe throw in a bit of Lesson 1, and explain, "And that is why we are not even going to bother with the rest of this unit."

What is the benefit of using a bot to do most of the work of a writing assignment? Who knows. This unit is a solution in search of a problem, or more to the point, a product in search of a use case.