There's a whole wing of the edu-wonk world pre-occupied with "high-impact tutoring," whatever that is presumed to be. The field serves as a successor to the notion of "accelerated learning," the idea that learning gaps could be closed by teaching the students who were behind by teaching them more, faster, than the other students-- an idea that stunk for a variety of reasons.
What all of these ideas have in common is that they are super-appealing to techbros anxious to sell some version of AI tutoring. Lauraine Langreo over at EdWeek took a look at some studies about the various ways AI can be deployed for tutoring, and she opens with what has been the guiding principle for all of these operations.
"Has been heralded" is doing a lot of work here, but the idea that it takes a lot of humans spending a lot of time (and wanting to be paid for it) is spot on. That crucial fact about tutoring smells a lot like money to AI companies. "However," as she notes, "research on the effectiveness of AI tutoring models is limited and still emerging."
Well, that's kind. Sal Khan has taken a good hard look at his own products and discovered that-- surprise--when you plunk a young human in front of a screen and say, "Go ahead and use this program to teach yourself math," by far the most common response is "I wonder if I can get regular Youtube on this thing." Students simply don't use it.
SCALE, part of Stanford's Accelerator for Learning (yes, that again), has a brief out that looks at the ways AI can be used, and well...tutoring by a human has a "robust evidence base." After that, every evidence base is "emergent," a research term that means roughly "there is no actual evidence of this in sight, but we kind of smell the possible concept of a piece of the beginning of a sprig of evidence just about to appear, but it's not here yet."
SCALE likes a human tutor with AI supporting the tutor behind the scenes, which they call "potentially as effective, or more effective" as an actual unaided human. This would, of course, depend on the human. But if Human plus AI gets you the same results as Human alone, that suggests at the very least that the extra cost of AI is a waste of money. If the problem your leaders are trying to solve is "How do we incorporate AI into our program," maybe this looks not so bad, but if the problem you're trying to solve is "How do we best tutor these young humans," then adding AI seems like a non-starter (even before we start considering, as SCALE does not, the various negatives associated with AI.
"AI tutoring with human support," which has been the model of computer-powered schools, from Summit to Altschool to Alpha School-- the emergent evident here suggests that implementation matters, because it's only a shade removed from AI-only tutoring, which SCALE declares has "unknown effectiveness." Sad news for everyone who thought they could just swap out an actual human tutor for some AI program and get "high impact" results.
Look, we know the most effective means of tutoring-- one teacher, one or two young humans, repeated a few times a week. It's not a mystery. But there are two longstanding responses to this. One is the eternal education response from leaders and taxpayers: "That sounds nice and I can see it would make education better, but I don't want to pay for it." The other is corporations looking at it and thinking, "I bet we could make money pretending to fix this problem."
But if you want to get some actual tutoring done, you are going to have to hire a knowledgeable human being to do it, not a bot.
At last we are seeing some of the press that Alpha School actually deserves. Lexi Churchill, in a joint Texas Tribune and ProPublica piece, shines a spotlight on the Texas emperor's wardrobe. You should absolutely read the article, but I want to underline and expand on a couple of points.
First, the switcheroo. As it has in every other state (except, of course, Arizona), Alpha failed to get approval for its attempted charter school launch in Texas. That's Texas, where the state board of education crew is not known for its rigorous standards. This is not a surprise; Alpha School's educational bona fides are still limited to puffed up unproven baloney. As Churchill notes, several journalists have now tried to verify Alpha's claims of magical success with just two hours of screen time a day, and so far, evidence has proven elusive.
But to get into the Texas voucher program as a vendor, a school just needs approval from the state comptroller's office. For that approval, a school need only show that it operates in Texas, that it’s accredited by an approved organization, that it’s been in operation at least two years and that it administers annual standardized tests. Alignment with state education standards? Proven student success? Evidence-based program administered by qualified personnel? None of that is needed.
It's worth remembering that Alpha is by no means the only school playing this game of switching back and forth between charter and voucher status based on whichever category will let them keep the money flowing in with the least possible accountability. If your state has both charter schools and vouchers, I guarantee that there are schools picking whichever designation will keep their owners happiest.
And if you wonder what exactly is the aim of Alpha school and the husband and wife team of MacKenzie and Andrew Price, there's this reminder.
At least two state education boards also took issue with Alpha’s plan to keep Price and her husband on the charter’s oversight board while the school simultaneously paid their companies, which provide services like the two-hour learning platform. The Prices acknowledged the conflict of interest and repeatedly promised to correct it.
There's Unbound Academy, which is the charter arm of Alpha School itself. Both tout the 2 Hour Learning program, which is set up as separate business. The proposal in the PA Unbound application said “In support of its operations, Unbound Academy will collaborate with 2hr Learning, Inc. to deliver its adaptive learning platform, while Trilogy Enterprises will manage financial services, and Crossover Markets, Inc. will assist with recruiting qualified virtual educators."
Trilogy Software is owned by Joe Liemandt, the billionaire techbro who backed all of this and is currently the principal of Alpha School; Andrew Price is the CFO. Crossover (an internet employment hub currently advertising a job as Director of Academics fore 2 Hour Learning at $400K) is supported by ESW, Joe Liemandt's personal investment company; their main client is Trilogy. Crossover has a proprietary tool called WorkSmart, which surveils employee work on their computer and was promised for the PA Unbound school. Crossover Markets and 2 Hour Learning share the same Austin address. It appears that Andrew Price was also linked to Crossover, but I find no clear current evidence.
The PA application also noted that the general and administrative service provider would be YYYYY.LLC, a company whose director, president, and principal are Andrew Price.
So Alpha School comes with a whole big web of companies that it "hires" to run the various aspects of the operation, and which in turn funnel all of that money back to the Prices. As Dan Meyer noted regarding that Arizona Unbound charter, "Ultimately, MacKenzie Price, Unbound Academy’s board president, will need to decide whether MacKenzie Price, founder of 2 Hour Learning, Inc, has set a fair price for those curriculum and resource materials. Arizona taxpayers await her decision."
Also worth noting-- Alpha School has ridden a wave of buzz mostly generate by MacKenzie Price's willingness to plug her operation to any media outlet that was willing to be impressed. And yet, according to Churchill
None of the Alpha leaders named in this story, nor the company’s press office, agreed to interviews, nor did they answer most of the newsrooms’ questions, including about their test scores. An Alpha spokesperson provided an emailed response only to questions about previous reporting by other news outlets.
So, feeling a little shy suddenly.
But hey-- so much synergy. Synergy between Alpha's lazy charter school and Texas's lazy voucher rules. Synergy within the web of money-scrubbing corporations. Who cares about weaving a tangled web if it's a profitable one.
Alpha School, if you've somehow missed the noise, is a private school chain created by Texas businesswoman Mackenzie Price, her husband, and a real rich guy. It leans on the idea that a child's basic core education can be fully conveyed by a computer program in about two hours a day: Alpha School is directly connected to 2 Hour Learning, because like many private school grifts, the company is one of many companies, wheels within wheels, all handing money to each other, and through each other ultimately to the owners.
Alpha School is expensive-- as pricey as an ivy league prep school, but the buzz has been phenomenal and ongoing. If Price could do education one quarter as well as she does PR, the school might actually have something to offer.
The Scientific American piece is notable for some of the worst arguments I've seen made for Alpha School, by the guy currently employed as the "senior learning scientist" for Alpha-- Carl Hendrick, who really ought to know better.
Carl Hendrick is an internationally recognized science of learning and instructional design expert. I'm about to favorably review a book he co-wrote. I have on occasion forwarded some of his substack offerings. People I respect, respect him.
And yet, the Carl Hendrick in this piece is really out in the weeds.
First, he compares AI in education to Waymo on the road. See, the autonomous car has to learn to accommodate wobbly cyclists and kids running into the street and non-straight roads-- all the things that Waymo, with its limited training for very specific urban areas, still has trouble doing.Sometimes big trouble. But Hendrick predicts that in "five or ten years" AI will be trained on "learning and all of kids' misconceptions and problems."
Thus waving away one of the problems with AI instruction-- the AI can't analyze where the student has gone wrong, what the glitch in their thinking or process might be. Is Hendrick suggesting that AI can programmed with a menu of every possible student misconception and then programmed with a response to it. Because that seems... unlikely. What he's proposing is that the program need not understand why the student is doing what they're doing, just like Waymo doesn't need to know why the bicyclist is wobbling. It's a strictly behavioral approach-- we need the kid to spit out the correct response.
Randolph slips past some issues. "When used by a teacher, adaptive learning models like Alpha School’s have shown promise in improving student performance and understanding." That sentence is doing so much heavy lifting. Used by a teacher? Models like Alpha School's? Promise? And she cites some objections from some academics. Talking to Kelly Miller, the Harvard professor who has worked in this area. Miller likes adaptive learning, calls it "the future of education." But:
But Miller uses AI and recorded lectures to cover basic material so that her in-person time with students can be spent on more complex work. If there’s no teacher-student relationship—like at Alpha School, where adults supervising students generally aren’t trained educators—“what’s the point?” she says.
Then it's all downhill from there.
Hendrick admits that Alpha School's model isn't perfect, "[b]ut he argues that most kids aren’t getting a good teacher or instruction grounded in learning science in a classroom anyway." So... their human teacher is probably crappy, so why not give them a crappy AI program instead? Hendrick argues that people would be shocked at the varying degrees of quality within schools. In other words, AI "educators" may be bad, but humans are worse-- or at least very inconsistent.
That’s Hendrick’s biggest argument for Alpha School’s model as a solution: “You can’t scale good teaching,” he says. Software, Hendrick argues, can be updated across classrooms at once as new curriculum and instructional research emerges and can give feedback far faster than a teacher grading assignments one by one.
Emphasis mine. Added while picking my jaw off the floor. You can't scale good teaching-- so let's scale mediocre-at-best teaching? And when it comes to feedback for students, which is more important-- being fast or being good? "I got feedback on my essay instantly. It's kind of dumb and useless and even wrong, but at least I got it back quickly."
Randolph also allows one quote that questions the selection bias involved (families that fork over that kind of money for private school tend to be highly motivated).
She also notes that the model will be tested in some public schools, which may show if the 2 Hour AI model works with a "broader" assortment of students, but we already have some data about that. But as this piece from The Conversation points out, much of the "data" we supposedly have is bunk. And as Dan Meyer showed, 2 Hour Learning has already been tested on general population students-- and it failed hard.
The irony here is that we are not talking about a revolutionary new technology-- the idea of a teaching machine that can guide students to cough up proper responses is over a century old. Read Audrey Watters Teaching Machines. Alpha School is parked firmly within a forest of red flags, and guys who are learning experts ought to be able to see them all waving.
McMahon's particular brand of passionless functionary was on display Sunday morning for a dozen minutes with Dana Bash. The interview kicked around a few topics, and featured one moment that highlight's McMahon's lack of understanding. I've watched it so you don't have to: here are the topics of the day.
The cost of going back to school. Bash opened with the cost of supplying a child for school. $175 for school supplies! "Packing a school lunch is up more than 10%," said Bash. "What do you say to parents" struggling with the costs and economic strains?
Good soldier and millionaire McMahon sticks with the party line of denial. There may be some price "upticks," but "the general economy is strong" and doing so well. Also, parents have a lot to look forward to because the administration has done work "to make sure that parents are more in charge of the education that is going on in the states" etc etc. So McMahon offers parents denial followed by changing the subject. Which leads directly to the department's hot new talking point--
Sex stuff. Expect to hear repeats of the statistic that one in ten students will have been sexually molested in school. McMahon manages to suggest that faculty are the main problem here and the administration is really working on that. She singles out California as place that shuffles teachers. Bash interrupts to ask if McMahon just pulled that stat out of her butt and McMahon says no the department did a study, based on some data-- "the general education statistics that were afforded to us." It should be pointed out that the California issue is a real thing, as reported by KQED and ProPublica, and the whole subject deserves attention (though this administration may not be the folks to do it). She says they have "really been cracking down on that as well" though how, exactly, is unclear. But it's two minutes into the interview and you already forgot that this started with a question about the high cost of sending a kid to school.
But God bless her, Bash did not forget and responded that sure some economic indicators are better "but the affordability crisis is real." McMahon offers a kind of half nod which could be a grudging acknowledgement of the point or a simple "Well, played, Dana."
AI in the classroom. Bash notes that McMahon issued guidance saying schools should evaluate AI tools based on evidence of effectiveness and learning outcomes. But, notes Bash, the tools haven't been around long enough for any such evidence to have been collected, yet Google and OpenAI are already shoving tools toward students. "Are you concerned that students are effectively becoming sort of guinea pigs for chatbots?"
McMahon (who has learned the difference between AI and A1) goes with the talking point that says we should watch out for screen time but make our caution based on good screens and bad screens. This launches her tale of going to Austin to see Alpha School in action. She was, of course, impressed.
The first couple of hours every day, students really have on headphones, they're in front of their computers, they're getting their subjects fed to them
Yup, that's how education works-- you just pour the knowledge into the brains of young humans and voila!! Instant smartitude. McMahon talks about "teachers" at Alpha School monitoring what's going on, but Alpha School doesn't use actual certified teachers for the job of "coach." McMahon says the beauty of the 2Hour Learning model is that "a student almost has a one-on-one tutor." They do not. AI is not a tutor for so very many reasons, not the least of which is that the AI cannot analyze where the student went wrong and guide them past that bump in the learning road. She describes how the teacher can see if the child is behind and therefore "repeat information" or see the child is ahead and speed things up-- all of which is what the magical AI is supposed to be doing, so I'm pretty sure McMahon comprehends zero parts of this whole set-up.
McMahon says with AI we should look at "who's going to get it, how long are students going to be exposed to it, what is the outcome," and that would be great for Alpha School, which carefully selects its students. We know, by comparing Alpha School outcomes with the outcomes of Unbound, their bargain cyber-charters that don't cream students, that their outcomes suck (see this invaluable Dan Meyer post).
McMahon agrees that AI products should be checked for effectiveness (not that she imagines a role for her department in doing so, because the MAGA concept of federal oversight is that there should be none and let the chips fall on those who deserve to be chip-burdened, so who exactly is checking these program for effectiveness or doing the pulling out is important, but not clear here). Anyway, she thinks Alpha School is so shiny, she supports "AI with guardrails," and again-- who is putting up these guardrails?
Bash asks if she thinks Alpha School is a proper model and she says, "Well, the PR pitch I received was convincing." I'm paraphrasing. She does note that it seemed effective at these schools for these students and she's not saying it would work everywhere, which is smart because we already know it doesn't.
AI opt out. Bash asks if parents should be able to opt out of AI, and the TL:DR answer is, "Yes." The school should make its case.
AI versus human. Bash kind of fumbles this question, but the upshot is "Is the swellness of AI tutoring outweighed by the loss of human interaction with teachers?" Even with a coach there, it's not the same as learning from a human being. Are you concerned about that. The TL:DR answer is, "No, that's not our problem." But she'll also allow that human vs. AI instruction requires a balance, as with any other new tool. Nothing can replace human interaction, both between students and teacher and between student and student. She's not wrong.
School phone bans. Does McMahon think they're a good idea. "Bell to bell no cell," she says. She's a fan. Then she's back to screen time balance, recreational vs. tutoring. This is a good a place as any to mention that McMahon was born in 1948.
CTE and College. Bash gets CTE education, but wonders if pushing career tracks in elementary school is a bit much. Of course, in MAGA world, where Certain People are destined for meat widget status and might as well get started early (see: every state that has weakened child labor laws). McMahon's quote here has some real layers:
I think what we can do is look at students. What are their interests? How do we direct them down a pathway that they're going to have an interest, you know, in school and I believe that children need a well-rounded education, but I think we have focused for too long for just that four year that path to a four year university and sometimes we ignore what the actual skill sets of those students are.
Note the idea that we are supposed to direct students down a pathway rather than, say, letting them choose a path on their own. Note also the notion that students' "skill sets" should determine their paths. I don't disagree that we have over-emphasized college, but the whiff of one of those systems where we sort students into future career paths early on, regardless of what the students themselves want-- well, that's a very stinky whiff.
McMahon cites middle school partnerships with community colleges, and Bash says, well, maybe that's debatable, but elementary school??!! Maybe get some critical thinking taught and leave specific trades till later. Nah, do both at once, says McMahon. But then she flips back to "we can't direct that from DC." Because each community and its school might "need" different things. This seems really close to the notion that poor communities need to teach students to be meat widgets, and wealthy communities don't. Certainly there is nothing in her formulation of education about providing opportunities or, ironically, actual choices for students. But she's right back to "and that's why we really want to return education to the states, because we shouldn't have to make sure students all have equal opportunities no matter their zip code." Okay, I added that last part.
Higher ed free speech. Bash brings up the letter McMahon sent to universities calling for more open admission and more free speech. Bash wants to know if the Trump administration will back that up with cuts of federal funds. McMahon's reply starts with "Well, we're certainly always going to monitor what our universities are doing" which is both creepy and also contrary to sending education back to the states, but she then cites an "incredible positive response" and argues that she's all about open debate by folks on both sides of the aisle. She cites Ron DeSantis saying that Florida is all about those open two-sided debates in higher ed, so clearly that's totally working.
Look, it's a foundational belief of MAGA that the dirty Lefties captured institutions, including higher education, and that to set the country aright, those institutions must be recaptured. Chris Rufo has talked about it repeatedly; somebody has a whole book coming out about it. That's the project. Anything else is a smokescreen. McMahon's call for a "collision of ideas" can't be taken seriously.
That's pretty much it. McMahon mostly snoozed her way through the interview. As usual, McMahon strikes me as someone who is perfect for the task of doing away with her job because it's a job that doesn't seem to interest her very much at all. Honestly, I'm not sure why she's here. Betsy DeVos wanted to bring kingdom rule, but McMahon is old and wealthy and seems like she's out of effs to give. She may want to watch the world burn, but I imagine she would not stick around long to bask in the glow.
You can watch it all if you want, but don't bother to make popcorn.
Several decades ago, my brother and I played in a strolling dixieland band at Conneaut Lake Park, a delightful small amusement park that has since fallen on difficult times, and one of the things we noticed at the time was that small children would "dance" to our music by running in little counterclockwise circles. Lo and behold, researchers have discovered that turning counterclockwise is an unexplained but real human thing. We humans truly are a mysterious species.
Here's your reading list for the week. Read it in whatever direction you like.
Julie Letchner provides a specific, local example of how one district confuses compliance with quality, and how full length books are kept out of the classroom.
Jan Resseger keeps track of Ohio education shenanigans, including the legislature's fondness for leaning into failed policies while refusing to support the public school system.
Not sure I've seen this take from anyone on any side. Frederick Hess asks why bother with education cuts if we're just going to blow a mountain of money and saddle the next generation with mega-debt?
Russell Frank, columnist for StateCollege.com, thinks maybe his AI devices are not doing great work. Best line:
The Machine can do a lot of things that we mere mortals cannot. But it doesn’t know what it doesn’t know, which means it may be artificially intelligent, but it isn’t artificially wise.
Nothing like a good rant. This rant by Brian Phillips is pretty delightful. Thanks to Benjamin Riley for highlighting this in his fine Punk is anti-AI post.
The Organization of American Historians has released a report that attempts to summarize all of the current administration's attempts to rewrite or erase history.
This week at Forbes.com I took yet another pass at explaining why federal school vouchers are bad news. It's not just the money-- it's the fundamental change to the public education mission. I'd be delighted if you shared this one with your favorite elected state official.
If you were a band kid in the early seventies, you listened to Maynard. We were lucky enough to see him live at Edinboro University every summer for a buck. When he scored a semi-hit with "Gonna Fly Now" that marked the end of MF Horn Maynard (concert closer: "Hey Jude") and the beginning of disco Maynard (concert closer: "Maria") but we didn't begrudge him his success, and later he moved back around to cool stuff like this:
Nobody likes to train teachers like Google, who, like tobacco companies, understand that if you can lock those customers in while they're young, you'll get to keep them for life. Hence, Google's pre-eminence in training teachers to use whatever damn thing they've come up with. The "pipeline of future users" memo was turned up by NBC journalist Tyler Kingkade. Kingkade just came back from a free two-day Google training camp for pushing Gemini, Google's house brand of odious AI, and the resulting piece is exactly the sort of thing I'm not supposed to read unless I wash it down with a double order of blood pressure medication.
But here, in one place, we can find so much of what is wrong about the AI-in-the-classroom boosterism. I've read this so you don't have to, but it's going to be neither short nor sweet. God bless Kingkade for suffering through the whole thing.
Kingkade opens with a group of k-12 educators "sitting in an atrium on Google's campus" trying to imagine what pushback they'd get from some old fossil of a colleague, the kind who is upset that cursive is no longer taught. She might "yell" that AI is just another shiny fad. "What’s next, she might ask — robots teaching kids how to read?" Oh, that whacky old fossil.
Where do they turn to get help with this dinosaur? Why, they ask Gemini for tips, of course.
They would win over this skeptical English teacher by explaining what generative AI could do for her: create classroom materials for phonics lessons, reducing what would normally be hours of work to just two minutes. The key, the educators agreed, was to avoid getting into an argument or letting this AI critic unload all her fears uninterrupted.
Got that? Don't listen to her-- don't even let her finish talking. Instead, act like a computer program, not a human colleague. And certainly don't treat her concerns as if they are legitimate. Focus on "pain points" says one leader, and how Gemini can take away your pain. And I have to tell you-- in my youth, I went through training for how to handle penitent peers who had answered an altar call, and some of this "training" seems very familiar.
“It’s not as scary if you’ve taken something off of my plate versus giving me a new thing that I have to then go out and learn,” Winston Roberts, director of an AI initiative at ISTE+ASCD, a nonprofit education group that worked with Google to develop the training, told educators from a stage.
When a non-profit is pushing a particular profit, you have to ask who is paying the bills. In this "partnership," somebody has invested a lot of their marketing budget. Teachers get way way wayyyyyy too much of this kind of "training" that is actually marketing masquerading as "help" for teachers.
Kingkade takes a moment to note the considerable growing backlash against AI in general and screens in classrooms in particular. Then we are back to the session.
We meet an English teacher from Hawaii who notes that AI is now part of student vocabulary; the example is that students use AI to call things fake, which ought to tell us a valuable lesson right there. And then she offers this comparison:
“If a student’s running in the hallway, you don’t take away the hallway — you teach them the proper behavior for the hallway.”
With all due respect to my Hawaiian colleague, no. If a student is shooting at things with a gun, you take away the gun. You don't let them drive a car until they are 16 and have passed requisite tests of competence.
Google rolled out it's "free" online training that includes "guidance about creating study guides, crafting lesson plans and analyzing where students are getting stuck," but not arguments for why AI might be a bad match for those tasks. Crafting lesson plans? That should be done by someone who knows the material and the students- AI knows neither. Analyzing where students get stuck? I'm unconvinced that AI can do that at all, but even if it could, the important question is WHY students get stuck. That is one of the most basic teacher functions-- working out what mistakes in thinking are taking the student into the weeds. Simple repeating, "Yeah, you are in the weeds again" is no help; students need help figuring out what wrong turn they took.
The training materials describe Gemini as “an engine for high-quality instruction” to do the “heavy lifting” for designing classroom lessons. “As an educator, this shift moves you into the role of a ‘learning conductor,’” one slide states.
Learning conductor? Learning conductor??!! Time for another blood pressure pill. Gemini is not an engine for high quality instruction because high quality instruction involves a human. Nor can Gemini design high quality lessons; it can only mimic and average the lessons it has input. And wait a minute-- isn't "learning conductor" a fancy update of "guide on the side," a version of teaching that is widely and justly mocked?
“It’s really, really important that we use it,” Joseph South, chief innovation officer for ISTE+ASCD, "because Google has bet a shit-ton of money on this, and we need these kids to grow up into paying customers." Okay, he didn't say that last part. “We can’t just ignore it, we can’t ban it, we can’t keep it out of our schools — that’s not gonna prepare us for the future.” That may be at least partly true, but "just lie back and let it roll over you" doesn't seem like great future prep, either.
Kingkade reminds us about the "pipeline" memo and reminds us that Google is company that convinced every that schools needed Chromebooks ("What if a laptop, but broken") which we're now thinking, maybe not. Kingkade has watched a 2018 presentation about keep Google atop education mountain includes this great quote
“Educators are sitting on a growing goldmine of data,” the presentation said, but they needed help organizing and making sense of it. If Google designed ways for schools to use student data, it would set “the stage for us to reinvent the education system through data.”
Goldmine of data indeed. Also, shades of our data overlords, who used to be so sure that if they had all the data, they could control the world. Remember the claim that given access to all the data, the company could tell the student what to eat for breakfast on test day? Or that students wouldn't have to take the SAT because we'd already know what they were going to get? Yeah, that company is now toast, but the notion that a "data-driven" ed system would be awesome (and profitable and provide a digital profile of future meat widgets that corporations would love)-- apparently that kind of creepy Big Brother thinking still has a home.
Speakers at the training emphasized that humans should always stay involved with any AI use and that technology shouldn’t replace teachers [or, you know, learning conductors]. They focused instead on how a teacher could use Gemini to create a comic strip that explains how greenhouse gases trap heat, for example, or how elementary school children could use AI to generate more realistic depictions of their ideas than they are capable of drawing.
What the hell is wrong with these people??! Yes, the most important part of a third grade project is how realistic the rendering of the art is! After these last few years, Google, with all of its millions, is still struggling to come up with a non-stupid use case for AI in a classroom. But hey-- let's have a heavy dose of irony --
Casey Cuny, a high school English teacher in a Los Angeles suburb, described asking his students to debate their takeaways from readings — like the concept of “doublethink” from “1984” by George Orwell — with Gemini before discussing them in class.
“It’s the best discourse I’ve seen in years on some of these Socratic seminars I’ve been running in my classes,” he said. “It does push the thinking when used intentionally and strategically. And remember that I’m still using teaching methods — I’m not just putting it on the AI and walking away.”
Yes, class, turn and talk to your surveillance plagiarism machine about themes in 1984. Time for another blood pressure pill. You know what else is good for pushing thinking intentionally and strategically?? Human teachers and students! But it's good to know that when he turns on the AI, he still stands right there. Many of these AI boosters remind me of the Common Core days when teachers would self-own with variations of "I couldn't do my damned job until I had the Common Core Standards to tell me what to do." Sigh. I'm sorry. Cuny is probably a lovely human being and maybe a fine teacher, but I cannot begin to describe the rage I would feel if I found out that this was what was happening in my child's classroom.
Kingkade notes that the indoctrinees knew they'd be facing "challenges in evangelizing for AI." But one tech teacher said the training equipped him to "show skeptics how AI could be beneficial to learning-- not just for cheating." And “They may not like it, but I don’t think that’s going to change things,” he said. “The naysayers are not going to stop it.”
Yeah, you can't do AI marketing without just insisting that it's inevitable, which is admittedly so much easier than trying to provide compelling reasons that a teacher with a free will and professional conscience should choose to incorporate it.
Google and ISTE+ASCD are planning a host of these trainings across the country so that teachers can learn that using Gemini is awesome and great, also, how to roll over those terrible old-timey teachers who want to question whether there's any actual good reason to hand the class reins to Gemini while they become learning conductors. Because "training" here means not teacher training, but unpaid field sales evangelist training. I'm going to go do some deep breathing exercises now.
As schools and teachers have tried to pressure their students to stay away from AI use, they have recapitulated many of the same old arguments against cheating in its traditional forms.
We English teachers have railed against shortcuts since CliffsNotes first reared their "study guide" heads back in 1958. Then the internet begat SparkNotes and its ilk. And it was always a mistake to frame the argument as some sort of moral or ethical issue. "You're a bad person if you cheat on this assignment," is not a useful message for young humans for many reasons, not the least of which is that they hear variations on "You're a bad person if..." a lot.
As it turns out, the best arguments against old school cheating are equally valid against new school high tech cheating, or just plain AI "augmentation."
Anything worth doing is worth doing yourself
"I would really like to kiss this highly engaging and exciting human being in front of me, so I am going to get someone else to do it and tell me what it was like," said nobody, ever.
You get the most out of life's experiences by, you know, experiencing them. You could sit in a cave somewhere and let your tech feed you a regular summary of what is going on outside, but what would be the point? You find your best self, you learn how to be fully human in the world, by being in the world.
Too many adults, and far too many adults who work in schools, feed the narrative that students are in some sort of holding pattern, that their real lives in the real world will start further down the road. That's just not true. Your life is going on right now, even if you are not yet an adult. So experience it first hand. And yes, that includes the work that you've been given to do in school.
Of course, "anything worth doing" is doing some heavy lifting here. That part falls on the teachers. It's part of their job to make sure they are bringing students together with things that are, in fact, worth doing; then they have the task of making the "worth doing" case to students.
Lying is corrosive
Everyone has seen the memo explaining that lying is wrong. But it's also important to understand that lying is corrosive and self-damaging. And it's nearly impossible to cheat without lying. And lying is corrosive.
Lying builds barriers in relationships; in particular, it ruins trust, and without trust as a foundation, it is difficult to build or sustain any sort of relationship with other human beings. Lying creates a brutal sort of isolation, in which you alone are the only person who knows the truth of your own story. That kind of isolation is the usual root of the whole existential angst thing anyway, but to add the barriers that come with lying just makes it so much worse.
As I told my students a gazillion times, life is too short to put your name to a lie.
Protect your brain
You do not build muscles by hiring someone else to lift weights in your name. Students are developing their minds, strengthening their brains. There is a natural tendency to draw back from the friction and pain involved, but that's how you build things.
Your brain is the toolbox that will hold every tool you'll need to make your way through the world, both personally and professionally. The more, better tools you collect, the more choices you will have in life. We know that offloading cognitive work to AI is not good for people. It's not good for adults and degrades the tools in their mental toolboxes, but for young humans who are supposed to be accumulating those tools the effects could be even worse-- the absence of necessary tools as they enter the adult world.
It is becoming increasingly clear that AI is not for amateurs, that it is only useful for people who are already knowledgeable about the field in question. Students are not those people.
You are going to need your brain your whole life, and your school years are the chance to pack it with as many bits of knowledge and skill you can get your mental mitts on. Do not use AI to shortchange that process.
This requires the kind of long term thinking that young humans does not always come easily to young humans. But we adults have to keep reminding them that the work is not to generate an assignment that you can hand in tomorrow, but to wrestle with the work in ways that will help them accumulate the knowledge and skills that will help them move through the world. Speedruns and shortcuts will not help with that.
Don't avoid cheating or cutting corners or just getting a little extra help because it's Very Naughty. Avoid all of these with either AI or old school methods, because they get in the way of the work of building your self and your life. That should your measure in all things-- is this a tool for helping you grow and live, or a means of avoiding engaging with growth and life? Don't choose the latter.
AI is not responsible for the rise if bad writing assessment, but it is promising to provide the next step in that little journey to hell.
Let me offer a quick recap of bad writing assessment, much of which I experienced first hand here in Pennsylvania. The Keystone State a few decades back launched the Pennsylvania System of School Assessment (PSSA) writing assessment. Assessing those essays from across the state was, at first, a pretty interesting undertaking-- the state selected a whole boatload of teachers, brought them to a hotel, and had them spend a weekend scoring those assessments.
I did it twice. It was pretty cool (and somewhere, I have a button the state gave us that says "I scored 800 times in Harrisburg). Much about it was not entirely impressive. Each essay was scored by two teachers, and for their scores to "count" they had to be identical or adjacent-- and on a five point scale, the odds are good that you'll meet that standard pretty easily. We were given a rubric and trained for a few hours in "holistic grading", and the rubric was pretty narrow and focused, but still left room for our professional judgment.
But then the state, like many others, stopped using teachers. It was easier to put an ad on craigslist, hire some minimum wage workers, train them for half a day, and turn them loose. Cheaper, and they didn't bring up silly things like whether or not a student's argument made sense or was based on actual true facts. (There is a great old article out there by someone who did this, but I can't find it on line).
Pennsylvania used this system for years, and my colleagues and I absolutely gamed it. We taught our students, when writing their test essays, to do a couple of simple things.
* Fill up the whole page. Write lots, even if what you're writing is repetitive and rambling.
* Use a couple of big words (I was fond of "plethora"). It does not matter whether you use them correctly or not.
*Write neatly (in those days the essays were handwritten).
* Repeat the prompt in your first sentence. Do it again at the end. Use five paragraphs.
Our proficiency rates were excellent, and they had absolutely nothing to do with our students' writing skills and everything to do gaming the system.
The advent of computer scoring of essays has simply extended the process, streamlining all of its worst qualities. And here comes the latest update on that front, from Tamara Tate, a researcher at University California, Irvine, and an associate director of her university’s Digital Learning Lab, her latest research-- "Can AI Prove Useful In Holistic Essay Scoring"-- written up by Jill Barshay in Hechinger.
The takeaway is simple-- in a fairly big batch of essays, ChatGPT was identical or within a point (on a six point scale) of human scorers (actual matching 40% of the time, compared to 50% for humans). This is not the first research to present this conclusion (though much previous "research came from companies trying to sell their robo-scorer), with some claims reaching the level of absurdity.
The criticism of this finding is the same one some of us have been expressing for years-- it says essentially that if we teach humans to score essays like a machine, it's not hard to get a machine to also score essays like a machine. This seems perfectly okay to people who think writing is just a mechanical business of delivering probable word strings. Take this defense of robo-grading from folks in Australia who got upset when Dr. Les Perelman (the giant in the field of robograding debunkery) pointed out their robograder was junk:
He rightly suggested that computers could not assess creativity, poetry, or irony, or the artistic use of writing. But again, if he had actually looked at the writing tasks given students on the ACARA prompts (or any standardized writing prompt), they do not ask for these aspects of writing—most are simply communication tasks.
Yes, their "defense" is that the test only wants bad-to-mediocre writing anyway, so what's the big deal?
The search for a good robogradcer has been ongoing and unsuccessful, and Barshay reports this piece of bad news.
Earlier versions of automated essay graders have had higher rates of accuracy. But they were expensive and time-consuming to create because scientists had to train the computer with hundreds of human-graded essays for each essay question. That’s economically feasible only in limited situations, such as for a standardized test, where thousands of students answer the same essay question.
So, the industry will be trying to cut corners because it's too expensive to do the job even sort of well-ish.
Tate suggests that teachers could "train" ChatGPT on some sample essays, but would that not create the effect of requiring students to try to come close to those samples? One of Perelman's regular tests has been to feed a robograder big word nonsense, which frequently gets top scores. Tate says she hasn't seen ChatGPT do that; she does not say that she's given it a try.
And Tate says that ChatGPT can't be gamed. But then later, Barshay writes:
The next step in Tate’s research is to study whether student writing improves after having an essay graded by ChatGPT. She’d like teachers to try using ChatGPT to score a first draft and then see if it encourages revisions, which are critical for improving writing. Tate thinks teachers could make it “almost like a game: how do I get my score up?”
Yeah, that sounds like gaming the system to me.
Tate has some other odd observations, like the idea that "some students are too scared to show their writing to a teacher until it's in decent shape," a problem more easily solved by requiring them to turn in a rough draft than by running it by ChatGPT.
There are bigger questions here, really big ones, like what happens to a student's writing process when they know that their "audience" is computer software? What does it mean when we undo the fundamental function of writing, which is to communicate our thoughts and feelings to other human beings? If your piece of writing is not going to have a human audience, what's the point? Practice? No, because if you practice stringing words together for a computer, you aren't practicing writing, you're practicing some other kind of performative nonsense.
As I said at the outset, the emphasis on performative nonsense is not new. There have always been teachers who don't like teaching writing because it's squishy and subjective and personal-- there is not, and never will be, a Science of Writing--plus it takes time to grade essays. I was in the classroom for 39 years--you don't have to tell me how time-consuming and grueling it is. There will always be a market for performative nonsense with bells and whistles and seeming-objective measurements, and the rise of standardized testing has only expanded that market.
But it's wrong. It's wrong to task young humans with the goal of satisfying a computer program with their probable word strings. And the rise of robograders via large language models just brings us closer to a future that Barshay hints at in her final line:
That does give me hope, but I’m also worried that kids will just ask ChatGPT to write the whole essay for them in the first place.
Well, of course they will. If a real human isn't going to bother to read it, why should a real human bother to write it, and so we slide into the kafkaesque future in which students and teachers sit silently while ChatGPT passes essays back and forth between output and input in an endless, meaningless loop.
If you'd like to read more about this issue, just type "Perelman" into the blog's search bar above.
"Hmm. Can you explain to me why you took this step here...?"
"That's an interesting interpretation, but I think you might have overlooked this..."
There are so many ways in which generative language algorithms (marketed as AI) can't do the work of a teacher, some larger than other.
Some are pretty basic. The notion that AI can create lesson plans only makes sense if you think a good way to do lesson plans would be to have an assistant google the topic and then create a sort-of-summary of what they found.
But other obstacles are fairly huge.
Certainly there's a version of teaching that looks like this:
Student: Here's an answer.
Teacher: That's wrong. Try again.
Student: How about this?
Teacher: Still wrong. Try again.
For different sorts of content, there's a version like this.
Teacher: Do A, then B, then C, and you will get X.
Student: Um, I got Q somehow.
Teacher: Do A, then B, then C, and you will get X.
Student: I'm not so sure about the B part. Also, I got V this time.
Teacher: Do A, then B, then C, and you will get X.
The technical term for this kind of teaching is "poor" or even "bad." Also, "teaching via Khan Academy." This also applies to new AI-powered versions like Khanmigo, which tries to help by essentially directing you to a video that specifically shows you B. Or you can throw in "special interests" and the AI will "incorporate" references to your favorite hobby.
Part of the work is to try to get inside the students' head. It is not enough to assess whether the student has produced an answer that is right or wrong or sort-of-right, and it's certainly not enough to repeat some version of "Don't be wrong. Be right," over and over again. The job is to figure out where they may have stumbled, to see where they are in the vast territory of content and skills that we are helping them navigate.
Part of the work is watching students struggle, watching the cues that they have hit a rough spot, collecting data that reveals how they are trying to work their way through the material, sorting and sifting the clues into important and unimportant sets. Part of the work is thinking about how the students are thinking. Part of the work is looking at how certain soft intangibles (e.g. the Habits of Mind) play out as the student wrestles with the material.
Sure, the algorithm can "learn" cues that indicate certain mistakes in thinking (if you looked at 2 x 3 and got 5, you probably added instead of multiplied), but the more complex the task, the more varied the outcomes, and the more varied and unpredictable the outcome, the less capable AI is of dealing with it (e.g. how does Hamlet's character arc reflect his relationship with death). Does the student show an attempt to really come to grips with the materials, or are they just spitting out something that the AI would recognize as correct?
It all makes a difference. Is the student soooooo close, or just flailing blindly? Is the student really trying, or just coasting? Is the student making operational errors, or operating with flawed fundamentals? Part of the work is to try to assess the student process, but AI can only deal with the result, and the result that the student produces is often the least critical part of the learning process.
For young humans, the best learning requires relationship with another actual human being. Eventually humans learn to teach themselves, but that comes later. Until that day, small humans need interaction with another human, some being who can do more than simply present the "right" answer over and over again.
