Against AI Theft

Among the many reasons to give Artificial Intelligence some real side-eye is the business model that rests entirely on plagiarism-- stealing the works of human creators to "train" the systems. Now a new paper attacks a defense of the AI cyber-theft machines.

"Generative AI's Illusory Case for Fair Use" comes from Jacqueline Charlesworth (who appears to be a real person, a necessary check that we all need to do now whenever we come across scholarly work because AI is cranking out sludge swiftly). Charlesworth was a general counsel of the US Copyright Office and specializes in copyright litigation. 

The folks hoping to make bank on AI insist that piracy is not their business model, and one of their favorite arguments to hide behind is Fair Use. Teachers are familiar with Fair Use rules, which tell us that we can show movies if they are being used for legitimate teaching stuff but not for entertainment. 

But as Charlesworth explains it, the Big Boys of AI argue that while the programs are copying the wo4rks used for training, the AI only "learns" uncopyrightable information about the works. 

Once trained, they say, the model does not comprise or make use of the content of the training works. As such, they contend, the copying is a fair use under U.S. law.

That, says Charlesworth, is bunk.

The 42 page paper combines hard-to-understand AI stuff with hard-to-understand law stuff. But it includes lots of useful insights and illustrations of AI's lack of smartitude. And Charlesworth is a clear and incisive writer. And she dismantles the defense used by Big AI companies pretty thoroughly.

Despite wide employment of anthropomorphic terms to describe their behavior, AI machines do not learn or reason as humans do. They do not “know” anything independently of the works on which they are trained, so their output is a function of the copied materials. Large language models, or LLMs, are trained by breaking textual works down into small segments, or “tokens” (typically individual words or parts of words) and converting the tokens into vectors—numerical representations of the tokens and where they appear in relation to other tokens in the text. The training works thus do not disappear, as claimed, but are encoded, token by token, into the model and relied upon to generate output.

Furthermore, the earlier cases don't fit the current situation as far as business aspects go-

The exploitation of copied works for their intrinsic expressive value sharply distinguishes AI copying from that at issue in the technological fair use cases relied upon by AI’s fair use advocates. In these earlier cases, the determination of fair use turned on the fact that the alleged infringer was not seeking to capitalize on expressive content—exactly the opposite of generative AI.

Charlesworth also notes that in the end, these companies fall back on the claim of their "overwhelming need to ingest massive amounts of copyrighted material without permission from or payment to rightsholders." In other words, "Please let us steal this stuff because we really, really need to steal this stuff to make a big mountain of money."

Charlesworth does a good job of puncturing the attempts to anthropomorphize AI, when, in fact, AI is not "smart" at all. 

Unlike humans, AI models “do not possess the ability to perform accurately in situations not encountered in their training.” They “recite rather than imagine.” A group of AI researchers has shown, for instance, that a model trained on materials that say “A is B” does not reason from that knowledge, as a human would, to produce output that states the reverse, that B is A. To borrow one of the researchers’ examples, a model trained on materials that say Valentina Tereshkova was the first woman to travel in space may respond to the query, “Who was Valentina Tereshkova?” with “The first woman to travel in space.” But asked, “Who was the first woman to travel in space?,” it is unable to come up with the answer. Based on experiments in this area, the research team concluded that large language models suffer from “a basic inability to generalize beyond the training data.”

Charlesworth gets into another area-- the ability of AI to reconstruct the data it was trained on. One of her examples is one that shows up in the New York Times lawsuit against OpenAI, in which, with just a little prompting, ChatGPT was able to "regurgitate" nine paragraphs verbatim of a NYT article. This ability isn't one we often seen demonstrated (certainly it is not in OpenAI's interest to show it off), but it certainly creates a problem for the Fair Use argument. They may not have a copy of the copyrighted work stored, but they can pull one up any time they want.

And she notes that the cases cited in defense are essentially different:

Pointing to a handful of technology-driven fair use cases, AI companies and their advocates claim that large-scale reproduction of copyrighted works to develop and populate AI systems constitutes a fair use of those works. But Google Books, HathiTrust, Sega and other key precedents relied upon by AI companies to defend their unlicensed copying—mainly Kelly v. Arriba Soft Corp., Perfect 10, Inc. v. Amazon.com, Inc., A.V. v. iParadigms, LLC (“iParadigms”), Sony Computer Entertainment, Inc. v. Connectix Corp. (“Sony Computer”) and Google, LLC v. Oracle America, Inc. (“Oracle”)—are all in a different category with respect to fair use. That is because these cases were concerned with functional rather than expressive uses of copied works. The copying challenged in each was to enable a technical capability such as search functionality or software interoperability. By contrast, copying by AI companies serves to enable exploitation of protected expression.

There's lots more, and her 42 pages include 237 footnotes. It's not a light read. But it is a powerful argument against the wholesale plagiarism fueling the AI revolution. It remains for the courts to decide just how convincing the argument is. But if you're trying to bone up on this stuff, this article is a useful read.

Supporting Teachers

How do we help teachers be better?

Robert Pondiscio is fond of pointing out that A) you go into battle with the army you have, not the one you wish you had and B) with around 4 million teachers in the country, many are likely to be just regular human beings and not super-teachers.

He's not wrong. On the national, state, local, and building level, the teacher corps includes some very excellent teachers, some pretty good ones, some average ones, and a small but non-zero number of not very good ones. (My argument remains that it's not a simple bell curve because the majority of people who would be bad at it either never try or give up fairly quickly.)

Policy and reformy folks have tried to deal with this distribution in a number of unproductive manners.

Firing our way to excellence.

A favorite with the technocrat crowd. this was the plan whereby test scores soaked in VAM sauce was going to create hard data that could be used to make hiring and firing decisions (or, in some cases, merit pay decisions). But sure-- if we just fire all the terrible teachers, we'd be left with nothing but the good ones. 

There are numerous problems with this, starting with the lack of a valid or reliable way to evaluate teachers. The Big Standardized Test is its own kind of sham, but Value-Added Measures can only dream of someday working their way up to junk science status. 

Sardine Superteacher

The flip side of firing to excellence. This idea was to find the super-duper teachers and plunk them in classrooms with a couple hundred students. (There was a time when they also liked the idea of hooking the super-teachers up to computers, but COVID took some of the bloom off the distance learning rose.)

Rendering

Another idea was to take the Highly Effective Teachers and move them to the low-achieving schools. This idea lost traction on the slippery idea that teachers had to be convinced, somehow, to take the different job. More money? Sort of. Send a team to grab them, drop a hood over their head, and throw them in a van? Probably illegal. 

All of these have the same problem

It's not just that it's really difficult to quantify how good a particular teacher is. It's that teacher effectiveness is dependent on context and environment. A teacher who's effective with 20 students is not necessarily equally effective with 200. A teacher might be very effective with one type of student and not with another. And despite being the best work at being "professional," sometimes teachers bring their own lives into the classroom. Plus, what are we asking them to teach? I found that I was actually better with multiple preps in a day than just teaching the same thing, but some of my colleagues struggled with that kind of grind. And there's just the influence of time and experience; I taught for 39 years, and I was not the same teacher every one of those years.

On top of that, teaching involves a teacher and a student, and that connection is also variable. Pick any teacher in your local school, one that you are certain is terrible, and I guarantee you that we can find students who will praise that teacher to high heaven. Likewise, pick someone known widely as a wonderful teachers; we can find students who will tell you how awful they are. 

I'm not going to argue that judging teacher quality is impossible. I am going to say that it is heavily influenced by context and environment and factors that shift regularly, making it hugely difficult to quantify teacher quality in such a way that the measures can safely and accurately be used to make major decisions about teaching careers. 

And even if you could...

What are you going to do? Fire a bunch of teachers and replace them with...? How much more practical is it to take the folks you have and help them be the best they can be. Will there be a non-zero number of non-salvageable teachers who have to be shown the door? Certainly. But can you direct (or re-direct) staff to be better? I think that's not only possible, but necessary.

So how?

We've seen bad ideas about this, as well. 

Carrots and sticks and sticks and sticks

There's a whole family of reformy ideas that starts from the premise that teachers know how to get the high-achieving results that policy makers want, but those teachers have been keeping the secrets of Teaching Well locked in their filing cabinets, waiting to be either bribed or threatened into finally unleashing all the awesome.

It's a premise that is both insulting and myopic. The vast majority of teachers are doing the best they can with the tools they have. But some reformsters (looking at you, Arne Duncan) treat teachers like the main obstacle to educating children instead of the people who are actually in classrooms trying to get the work done. 

Teachers and their students face a variety of obstacles, but reformsters got into the nasty habit of dismissing these explanations of real hurdles by calling them "excuses" rather than, say, "challenges that maybe we could try helping teachers meet." 

De-professionalizing the Profession

For some reformsters, the dream has been the teacher-proof classroom. Set out a curriculum so specific and pre-programed that whoever your teacher is, you just hand them the program, tell them to implement it With Fidelity, and voila!-- an educational program that a trained monkey could implement effectively. Maybe it's scripted. Maybe it's just a day by day, minute by minute guide. Maybe it's a computer program, or a series of videos. 

Some reformsters see this as an opportunity to cut personnel costs. Turn teaching into a job that anybody can do, as long as they follow instructions, means that the labor pool is huge and the meat widgets hired for the classroom can be easily (and inexpensively) replaced. These are the folks who are so excited about AI "teachers" that they barely bother to pretend that such a move would foster better teaching. 

De-professionalizing teaching is the fast food model of education. But the promise of standardization in a McDonalds is not that you can always get excellent food there, but just the promise that you probably won't get terrible food. People who want excellent food go somewhere else, where the chefs are chefs and not assembly-line food prep meat widgets.

Maybe forcing your less-gifted teachers into a program to implement With Fidelity will improve schools on the bottom end (though I'm not convinced that someone with limited teaching ability can really implement one of these scripted programs effectively). But by stripping teachers of autonomy, you will inevitably hamstring your best teachers. You can argue that their superior teaching skills will allow them to find ways to put their personal spin on the mandatory teacher-proof program, but I'd say you're just arguing that they can still be great to extent that they work around, ignore, and otherwise find ways to escape the mandatory program. 

Removing teacher authority may or may not help your mediocre teachers, will hamstring your better teachers, and will make the profession less attractive to people who would be a real asset.

Testocracy

We are already reaping the problems created by a new generation of teachers who have never known anything in school except test-centered prep work. Too many have learned that you check the standards, google for "exercises" aimed at those standards, hand them out, drill them down, and that's supposed to be teaching (Daniel Koretz writes about this in The Testing Charade). 

Okay, have you got anything other than complaints?

The job of a school administrator (of any manager) is to create the conditions under which staff can do their best work. Most folks who work in schools already know that teachers are overworked, overstressed, and overburdened with a whole bunch of responsibilities. Many attempts to improve teaching and/or sell new education-flavored products are built around the idea that we could take X off teachers' plates. 

The crazy thing about this is that these attempts are all marked by one feature--the people behind them have decided on teachers' behalf what it is that teachers need. Let me suggest a crazy new approach--

Ask the teachers. 

I've talked about this before (see "The Seven Most Powerful Words in Education") but not enough since. Just ask staff, "What can I do to help you?" 

Now, sometimes this will be tricky, because teachers, as different individual human beings, will want/need different things. And sometimes they will want things admins can't give them. But administrators have to be better. One former colleague of mine pissed off our administration by asking, after being given yet another new responsibility, "What do you want me to stop doing so I have time to do this?" It's a legitimate question, and one that every teacher in that room was thinking, but asking it was Forbidden Not A Team Player naughtiness.

Ask your staff what they need. Don't just jump to "I have decided that what you want is a program to manage grades" or "I have decided that you want PD about apps I'm sure you want to use." Ask.

Provide high quality materials and resources.

But do not mandate how they must be used. Involve staff in the collection or creation of these materials, and revisit the collection annually. There will be eternal debates about which materials are high quality, just like there are eternal debates about what belongs in the canon. These debates are eternal because the answer keeps changing because of the times, the context, the available materials. But the fact that these debates can't be settled conclusively is not a reason to abandon the work of getting the very best materials available to your staff. 

Provide structure and scope and sequence, but don't set it in cement.

Your newer teachers should be able to find a useful answer to "What should I be working on next" and your experienced teacher should be comfortable adjusting the scope and sequence to fit the class, and all staff should feel safe adding their own special educational touches.

Flexibility is a local thing

Note that every mandate that comes down from the state or federal government tends to reduce flexibility, particularly since so many of them are wrongheaded variations on "If we make all teachers do X, all students will learn Y," a statement which is always wrong, no matter what you plug in. Policy makers need to ask one simple question-- does the proposed policy provide support or a straightjacket?

Teacher training and peer support

Too many undergraduate teaching programs waste too much time. For secondary teachers, there should be far more emphasis on the content of the subject area they plan to teach. For all teachers, there should be far more support through the student teaching experience, and hefty support should also be present through the first couple of years in the classroom. 

Some folks like the mode school model, with professionals working their way up as interns etc etc. Schools don't have to look exactly like that, but supports need to be in place. In most schools, whether a teacher has good support in their first few years depends on random factors like which other teachers have lunch the same shift. That early mentoring needs to be deliberate, intentional, and carefully considered. 

Peer support should continue. It should be easy for teachers who work in the same department or who work with the same cohort of students to collaborate and consult. If policy makers want to encourage this, there is one thing they can offer--money. Schools don't build more deliberate mentoring programs because such programs depend on time during the work day which equals money (sometimes there's also a lingering attitude that teachers are only really working when they have students in front of them).

Nothing else-- not PD, not merit pay, not threats, not scripted instruction-- will work to turn a new teacher into a good teacher better than regular support and mentoring by capable colleagues. 

Hold teachers accountable

The myth that teachers are all about defending low-achieving teachers is bunk. Second only to parents, nobody is more bothered by a low-quality teacher than the teacher who has to teach those kids the following year. What teachers fear is not accountability, but random irrational bad-faith harassment and mistreatment trying to pass itself off as accountability. 

So when a teacher wanders into the weeds, go help them get back. Yes, maybe they won't be helped, but you need to try first because firing just means starting over from scratch (if you can even find somebody). Give them extra-intense mentoring, coaching, daily assistance--whatever you think will get them back on track. But don't just leave them out in the field flailing. 

The non-answer answer

There is no one single simple answer to finding and developing good teachers, but we have more than enough experience to know that "Hire some people and hope for the best" is not the winning approach. Provide and surround them with access to top quality materials. Provide them with personal support. Treat them like grownups. Provide a supporting structure that holds them up without choking them off. 

If you want a metaphor, here's one I'm sure I've used before. 

Let's call teaching the classroom version of playing jazz. To pay jazz, you need a couple of things. For one, you need a solid rhythm section; a solid rhythm section makes everyone else sound better, plus it gives you a foundation on which to play. On that foundation, you have plenty of freedom, but you exercise that freedom within a framework--a best, chord changes, maybe even the basic tune of the song. Ignore the beat and the chords at your peril; you can't just do whatever the hell you want. Find good people to play with, and you will play better. And when you are really good, you can actually bend and defy the framework of beat and chords--but you have to really know what you're doing. It's part inspiration and gut, but it also requires technical skill and control and a good piece of equipment on which to play. Also, some songs are way better to play on than others (depending on who's on the stand tonight), so have a big book to select from. And especially also, you have to pay attention to your audience and where your own playing is in that moment, and adjust accordingly.

I have no idea how many jazz trombone players there are in the US, but if there were 4 million, only a small number would be a Jack Teagarden or a George Brunis or a Gunhild Carling. But given the right tools and the right support, the rest could do a good job. That's teaching. 

ICYMI: Here Comes The Arctic Air (1/19)

 Here at the Institute we are hunkering down and preparing for a blast of arctic air over the next two days. The Board of Directors gets tomorrow off in honor of Martin Luther King, Jr., which is the only thing happening tomorrow that I expect to pay attention to. 

A reminder that you can always help amplify stuff by posting it throughout your various social media channels. As someone who's able to track activity through at least one of my outlets, I can tell you that one just never knows where a particular post or article will catch fire. Your share could make a big difference in how widely something is read. Help folks out and share their stuff.

I am also happy to get recommendations. I read a lot, but I don't read everything, and I don't always get everything I've read into this weekly digest. So suggestions are always welcome. I had originally dreamt that maybe the comments section of these posts would fill up with "You should also read--" comments, and that hasn't happened, but the dream still lives.

In the meantime, here's this week's list.

‘Their Kind of Indoctrination’

In the New York Review of Books, Diane Ravitch on the kinds of attacks public schools can expect under the new Trump regime.

3 myths about rural education that are holding students back

Awkward structure aside (the three items are truths that debunk the myths), this is a welcome look at a more accurate picture of rural education.

‘Bless his heart,’ says Pulaski superintendent after ‘school choice evangelist’ sues KY district

Corey DeAngelis is butthurt that a Kentucky superintendent blocked him from attacking the district for supporting the anti-voucher measure that Kentucky passed. The superintendent is not impressed.

On a Mission From God: Inside the Movement to Redirect Billions of Taxpayer Dollars to Private Religious Schools

Alec MacGillis looks at the story of how Ohio set out to get public money into Catholic private schools. Choice was just a tool. This is well-researched and detailed and a bit alarming.

Ohio’s Journey from Public Education to School Privatization

Jan Resseger, a retired Ohio educator, reacts to MacGillis's article.

Jeff Bezos Wants to Go to the Moon. Then, Public Education

Dominik Dresel at EdSurge and a convincingly scary look at Bezos and his long term plans for privatizing education.

The Uber Rich Are Funding “National School Choice Week” to Attack Public Schools

We'll all be hearing about School Choice Week soon, At Truthout, Alyssa Bowen, Ansev Demirhan, and Lisa Graves explain who's really behind it, and what they're after.

Volusia School Board member vows to stay despite Moms for Liberty chapter chair's threat

In Volusia County, Florida, the Moms for Liberty chair is opposing a former ally for being way too racist and insulting and mean. 

Defunding Public Schools is Really Unpopular

Jennifer Berkshire, blogging at The Education Wars, takes a trip to New Hampshire to watch democracy once again put the smackdown on an attempt to undermine public schools.

A new governor sets her agenda.

Also in New Hampshire, Andru Volinsky looks at the agenda of the state's new governor.

The Far Right’s Plan to Force Teachers to Lie About Race

Jesse Hagopian in The Nation outlining the threat of the Trump administration toward teaching a more authentic United States history.

I'm Not Sure Schools Can Teach Creativity

Can schools teach creativity as a sort of disembodied transferable skill? I don't think so, and neither does Chad Aldeman.

Measuring Artificial IQ

ChatGPT did a Thing, even a cool thing. But what does that mean, exactly? Benjamin Riley considers the question.

Top scholar says evidence for special education inclusion is ‘fundamentally flawed’

Is there any more reliable pendulum in education than the swing back and forth between putting students with special needs in regular classrooms vs. giving them a specialized separate room of their own? Jill Barshay at Hechinger reports on new research that will keep the debate going.

Are Today’s Students Really Less Independent Than Previous Generations?

At EdWeek, Arianna Prothero is really reporting about SEL program effectiveness in schools, which is also a topic worth discussing.

The MAGA Think Tank Behind Linda McMahon’s Education Agenda

Linda McMahon has been running a think tank that has served as a holding tank for Trump administration members waiting for their second chance. What that think tank has been saying may tell us what to expect from McMahon as Ed Secretary. Christopher Lewis and Jacob Plaza report for The Nation.

TIMSS Scores Down-- Don't Panic

Thomas Ultican digs into the latest in internation standardized math test scores. How bad are they, and do we really need to care?

Heroes, Hypocrisy, and Hubris

There's more on the ground detail here from TC Weber about Tennessee's new voucher push, but mainly there's a story about a teacher who has been put through hell and deserves to have his "not guilty" verdict published high and low.

Why Senate Cabinet Hearings are a Lesson for Schools

Nancy Flanagan and the problem of character and power.

Banned Book: Normal People

Steve Nuzum has been closely following the South Carolina committee charged with book banning for the entire state. Here he takes a close look at one particular book they chose to ban, searching for some hint of what their actual criteria might be.

AI Is Like Tinkerbell: It Only Works If We Believe in It

At Futurism, Jathan Sadowski suggests we think "of AI futurism as a sophisticated form of check kiting — cashing a check today and hoping the money will be in the account later." Predictions as marketing.

Know how to read cursive? The National Archives wants you

Really. If you are a master of this arcane art, the National Archives have tons of manuscripts they need to have translated into legible English. And you do it from home.

At Forbes.com this week, I looked at a new survey that shows, once again, people would rather fund public schools than vouchers. 

And an unusual week at the Bucks County Beacon, with two pieces-- a look at NPE's report on the massive failures of charter schools, and a piece about the attempt to launch an all-AI, no teacher cyber charter in Pennsylvania. 

You can also subscribe to my free newsletter and get all of my stuff in your email inbox. 

OH: Public Education Under Attack (Again)

The reporters at the Ohio Capital Journal have been all over this story. I'm here to give you the broad outlines of this latest adventure in trying to eclipse Florida and Arizona as the nation's leader in hostility to public education. 

The background: Ohio is yet another state that was ordered by the courts to fix its public education funding system (three decades ago). They've been trying to fix it with the Cupp-Patterson plan which shifts the burden from real estate taxes to the state government. 

As every state that has been through this (or has avoided going through this) knows, getting from inadequate funding to fully funding public education can cost a lot. In Ohio's case, that's about $2 billion over six years.

This makes some members of the Ohio GOP sad, especially House Speaker Matt Huffman, and the word "unsustainable" was thrown around. Huffman says that if they're spending that kind of money, why aren't public schools more accountable for how it is spent. Huffman told reporters that he thinks the final push of Cupp-Patterson just isn't going to happen, because of all that money.

As Morgan Trau reported for the Ohio Capital-Journal:

“That’s often how a lot of projects go — early on it doesn’t cost very [much] money — but some other governor or General Assembly will have to figure out how to pay for it,” [Huffman] continued. “As it turns out, I am the other General Assembly years in the future, or possibly am, and I don’t think the spending is sustainable.”

Just because some previous elected officials made a government commitment, that doesn't mean other elected officials have to honor it. Let's just cut those school funds.

The kicker here is that Huffman is also a huge supporter of Ohio's voucher program EdChoice, a program that has sucked up $1 billion taxpayer dollars and which involves no accountability.

So if the funding of public education is "unsustainable," how can the public funding of unaccountable private schools be sustainable?

Susan Tebben reported for the Capital-Journal:

” If the speaker thinks there isn’t enough education funding to go around, Ohio law is very clear,” Melissa Cropper, president of the Ohio Federation of Teachers, told the Capital Journal. “The legislature must fund public schools and make cuts to the costly and ineffective universal private school vouchers that were put in place by Speaker Huffman (as an Ohio senator) and other legislators,” said Melissa Cropper, president of the Ohio Federation of Teachers.

Those who support the funding model pointed to the $1 billion that went to scholarship funds including the EdChoice private school voucher program in 2023, which the legislature approved to give Ohio students near-universal eligibility to move to private schools of their choosing if they live in public school districts considered under-performing.

“If the speaker wants to talk about sustainability, you have to start with those numbers,” [Ohio Education Association President Scott] DiMauro said.

 Apparently plenty of legislators' phones have been ringing, because the Capital-Journal has heard from some of them. Reports Trau last Saturday:

Following our reporting on a proposal by Republican leadership in Ohio to cut public school spending, which resulted in the lawmakers facing backlash, half a dozen GOP legislators personally reached out, vowing to protect K-12 education.

Those six, and at least 15 others we have spoken to in recent weeks, say that one of their main priorities is supporting public schools.

Ohio has recently taken their vouchers universal, meaning that taxpayers now help pay off the tuition of wealthy families who were attending private schools already. Ohio is also a levy state where any tax increases by local schools have to go to a public vote, which means everyone gets to feel public school financial struggles immediately and acutely. 

As we learned this week from reporting by Alec MacGillis at ProPublica and the New Yorker, Ohio's school choice program have their roots in a desire to get taxpayer dollars to fund Catholic schools. The level of sneakiness and political gamesmanship that went into those maneuvers is kind of astonishing. It almost makes one appreciate Huffman's straightforward statement of his intention to throw public education under the bus while supporting the forced taxpayer support of unaccountable private schools. Here's hoping the actual taxpayers in Ohio don't let him get away with it. 

FL: Affirmative Action Revived for Some

You may recall Governor Ron DeSantis has a project to take over Florida's New College and turn it into a conservative bastion. Manny Diaz, Florida's unqualified education commissioner said their hope is "that New College of Florida will become Florida's classical college, more along the lines of a Hillsdale of the South.”

You may also recall the DeSantis is a staunch opponent of affirmative action, proudly telling the Moms for Liberty crowd that he goes even further than simply not doing AA. 

It's a standard MAGA talking point-- decisions should be merit only! They take considerable pleasure in quoting Martin Luther King Jr on being judged by content of character rather than color of skin.

But now it turns out that maybe a little affirmative action is okay-- as long as you're affirming the right thing.

New College hasn't just dismantled the gender studies department. That's old news. Steven Walker reported for the Sarasota Herald-Tribune about the new wave of students at New College, and, well, academic excellence does not seem to be their defining feature.

Overall, the average ACT and SAT scores for the incoming fall class at New College were lower than the previous year. The same group's overall GPA was also lower than in fall of 2022, according to data obtained by the Herald-Tribune and confirmed by the college.

Much of the drop in average scores can be attributed to incoming student-athletes who, despite scoring worse on average, have earned a disproportionate number of the school's $10,000-per-year merit-based scholarships.

 New College is recruiting heavily for athletes because Richard Corcoran, interim president and former head education privatizer for Florida, announced back last March that the school was going to build an athletics department. From scratch. Walker reports that of the 328 incoming students, 115 are student athletes. They scored 47% of the merit scholarships. Note: New College doesn't have athletic facilities, nor have they been accepted into the National Association of Intercollegiate Athletics.

They are also mostly men. Michelle Goldberg, writing for the New York Times, gets an explanatory quote from MAGA culture panic leader and New College trustee Chris Rufo:

In the past, about two-thirds of New College’s students were women. “This is a wildly out-of-balance student population, and it caused all sorts of cultural problems,” said Rufo. Having so many more women than men, he said, turned New College into “what many have called a social justice ghetto.” The new leadership, he said, is “rebalancing the ratio of students” in the hopes of ultimately achieving gender parity.

Well, heck. Even Reason, the very Libertarian magazine/website knows what that is. Emma Camp hit it under the headline "New College of Florida Embraces Affirmative Action for Men." They picked up the above quote, as well as noting that Rufo afformed that the whole point of expanding athletics was to draw more males. Rufo has written and ranted before about the "problem" of "the great feminization of the American university."

There are all sorts of issues here, but I'll let the Libertarian right call the college out on this:

But even if one accepts the premise that female-led institutions are more left-wing than male-led ones, New College's response—to engineer a more conservative institution by reducing female participation in it—is inherently in conflict with the strong defense of meritocracy and opposition to affirmative action espoused by Rufo and his allies.

Young women are outpacing their male counterparts in the academic sphere; girls now comprise roughly 60 percent of college students. While young men's failure to academically compete with their female classmates is worth our concern, the solution is not to lower the bar for men to obtain collegiate gender parity (something many selective colleges have been quietly doing for years). Nor is the solution to an illiberal left-wing campus culture to engineer a more male—and supposedly more conservative—student body.

Is the problem simply that MAGA doesn't like women very much? Sure seems like it some days. But for the moment let's just note that they are perfectly okay with affirmative action for dudes, which sure seems to support the theory that they were never against affirmative action--just the people they felt it was affirming, and the whole meritocracy argument was bad faith baloney.

But hey-- they are on the way to more closely resembling Hillsdale, with its student population of 865 men and 813 women. 

AI Can't Imagine Future Humanity

It's a minor throwaway article, but it is a fine example of how people who aren't paying close attention both accept and perpetuate huge misconceptions about what AI is or can do.

The headline from a story that originally ran on Tom's Guide, but was picked up by MSN (which is itself a bad sign) is "I used AI to imagine humanity in 50,000 years — here’s how it went."  The piece is by Ryan Morrison, the AI Editor for this tech-centered website. The first three paragraphs tell us how far into weeds we are headed.

I’ve always been a daydreamer, leaving my mind to ponder the possibilities of what could be to come. With the help of artificial intelligence tools, I can turn those ponderings into something I can actually see and even interact with.

Recently I found myself talking about space travel with ChatGPT, asking it about timelines and the impact terraforming a smaller world like Mars might have on human physiology. This later led to me having ChatGPT outline how it perceived humanity over 5,000, 10,000 and 50,000 years of evolution.

I also had it come up with ways humans might change if left isolated on different terraformed planets such as Mars, with its lower gravity or even the moons of the gas giants. I then used Freepik’s impressive Mystic 2.5 image model to bring them to life.

Morrison goes on to talk about how he offered ChatGPT different parameters and asked some other pointed questions. It all seems built around the notion that when he asks ChatGPT these questions, ChatGPT goes and looks at all the scientific research surrounding Mats and human physiology and gravity and whatnot and works up a series of theories based on a rational consideration of all the pertinent science. "Well, what if human civilization splits with no contact for a few millennia?" he asks, and ChatGPT strokes its chin and says, "Well, let me consult some sources and run some numbers." 

The article is laced with references to his "conversation" with ChatGPT. The program went on to "outline" how it "perceived" human change. He asked it to "imagine" humans in 50,000 years. 

Of course, ChatGPT doesn't do any of that. The stochastic parrot strings together an assortment of probable words in a probably string given the prompt, and given whatever training it has on sources that string together words near words similar to the prompt words. 

People like to think of "artificial intelligence" as the equivalent of some really smart, extraordinarily well read professorial type, or perhaps any of the artificial personalities we know from popular fiction. People who have AI products to sell like that picture of AI very much-- but Artificial Generalized Intelligence like that is not here yet, and may never get here. GPT-5 also not here. 

In the meantime, people who really ought to know better keep pretending that Large Language Models like ChatGPT are really something far more advanced (and useful) than they really are. But here's Morrison, saying his website bio has been written for him by ChatGPT, a "silicon-based life form."

This is the kind of stuff that trickles down to the general public and teachers and administrators and leads them to put all sorts of faith in "AI" that it does not deserve and cannot live up to. If we're going to have conversations about AI's proper place in the classroom, they will have to be based on reality and not marketing puffery and the imagination of over-excited commentators.

 

ICYMI: Fire and Ice Edition (1/12)

Whether you are being hit with a blizzard or a fire, I hope you are staying safe this weekend, and that you are coming through with minimal damage, and that you get the help and support that you need. For the rest of us, here's one researched list of places to which you can contribute to help folks in LA.

Here's some reading from the week.

Schoolhouse Crock

From The Baffler, Jennifer Berkshire's very excellent review of Adam Laats's very excellent book about one of the first great education con artists. 

Burning Down the Schools

Anne Lutz Fernandez and the impact of climate change on schools.

Writing Process: Scholarly/Academic Writing Edition

Paul Thomas shares practices and ideas surrounding the teaching of writing. He's an expert. 

Ryan Walters is blaming teachers for New Year’s attack. Has he forgotten Oklahoma history?

Ryan Walters, the education dudebro-in-chief of Oklahoma, continues to make it hard to believe that he was once a respected history teacher. But I guess if you have a Trumpian desire for press attention, you just have to keep saying stupid things loudly.

The Danger of Miseducation

Jess Piper connects Dylann Roof, January 6, and the problems that come with the rewriting of history.

Undoing EdTech's Death Grip on Education

At Restore Childhood, Denise Champney does some outstanding work breaking down just how deep and bad the edtech hold on education has become. 

Technology is supposed to decrease teacher burnout – but we found it can sometimes make it worse

"Use this app! It will save you time!" Here's some actual research to back up why every new piece of tech fills teachers with existential dread.

Everything Everywhere All at Once

Audrey Watters connects some techno-dots, starting the Power School data breach, in which somebody got their hands on a bunch of student data that the company was probably sell anyway.

The Ugly Truth About Vouchers

Paul Bowers at the ACLU explains why South Carolina's love for vouchers is just a bad, bad idea.

Fact-checking Elon Musk's claims that NJ teachers 'don't need to know how to read'

You may have heard President Musk's complaint that Jersey teachers will no longer need to be able to read. This piece from Lori Comstock explains why he's full of it, with all the details so you can explain it to your MAGA uncle.

Hundreds of Charter Schools Will Fail, Close, and Abandon Thousands of Families in 2025

Shawgi Tell looks at the research and lets us know what we can expect from charter schools this year.

AI Wants to Help Me Write– But With Disclaimers

The indispensable Mercedes Schneider tried taking an AI writer out for a spin, and while some results are predictable, take a look at the caveats that the AI company puts on its own product!

EnronAI

Back when he was a baby lawyer, Benjamin Riley worked for Enron. Yes, that Enron. He was on the inside for the early stages that led to that famous collapse (in a department trying to keep it from happening) and for him, much of the AI industry has a familiar smell.

Committee Moves to Ban More Books

In South Carolina, one committee can ban books from every school in the state. How efficient! Steve Nuzum reports on their most recent targets.

Newton Falls implementing Armed Staff Program

This is in Ohio, where a district is starting to arm its staff "to act as a deterrent and a force multiplier." Yeah, I'm sure that will work out.

Honor President Carter: Save and Improve the U.S. Department of Education!

Nancy Bailey says if you want to honor Carter, help protect and improve part of his legacy.

Mississippi Association of Educators Opposes Private ‘School Choice’ Efforts

Erica Jones, head of the MAE, will say it again-- don't strip funding from public schools to give it to private schools.

Trump’s Immigration Proposals Would Traumatize Children and Schools and Jeopardize Children’s Civil Rights

Jan Resseger dives into the question of what Trump's professed intent to throw out all the immigrants (that don't work for his friends) might do to their children. 

Is 2025 the Year to Eliminate Florida’s High School Exit Exams?

Florida is usually in the forefront of bad trends. Could they actually join the crowd on a good one? Sue Kingery Woltanski looks into it. 

Scandalling Up in Ohio

David Pepper profiles J. D. Vance's likely replacement, whose previous achievements include enabling one of Ohio's biggest privatized education scandals.

At Forbes this week, I explained why this is the heart of the school year, and looked at a scary bad new bill in Indiana for dissolving public school districts.

In the meantime, you can subscribe to my free substack and get all of my stuff in your email inbox.