OpenAI's Latest Snake Oil: Teaching Teachers How to Teach (Because They Clearly Don't Know How)

Nov. 20, 2024

Look, I’ve been staring at this press release for three hours now, nursing my fourth bourbon, and I still can’t believe what I’m reading. OpenAI - you know, those folks who brought us ChatGPT and a whole lot of existential dread - now want to teach teachers how to teach. Because apparently, that’s what education needs right now: another tech company mansplaining pedagogy to professionals.

They’ve rolled out this fancy “free” course (first hit’s always free, kids) in partnership with something called Common Sense Media. The irony of that name is so thick you could spread it on toast. Here’s the deal: it’s a one-hour, nine-module program designed to help K-12 teachers incorporate ChatGPT into their classrooms. Because what every underpaid, overworked teacher needs is another tech tool to master between grading papers and breaking up hallway fights.

According to their internal research - and boy, do I love internal research, it’s about as reliable as my ex-wife’s promises - 98% of participants found the program useful. That number’s so perfect it makes me suspicious. Last time I saw numbers that good, I was reading about North Korean election results.

But here’s where it gets interesting. They’ve already deployed this thing in places like the Agua Fria School District in Arizona. I’ve been to Arizona. It’s hot enough to fry an egg on the sidewalk, and apparently hot enough to fry common sense too.

The real kicker? Some educators aren’t buying this digital snake oil. Lance Warwick, a sports lecturer at UI Urbana-Champaign (who probably knows a thing or two about snake oil in the supplement industry), points out that the program contradicts itself faster than me trying to remember what happened last Saturday night. They’re telling teachers to input past grades and feedback in one breath, then warning them never to input student data in the next. Make up your mind, folks.

Then there’s this artist-educator from Arizona, Sin á Tres Souhaits (I had to down a shot just to pronounce that name), who compares AI to crypto. And damn if that isn’t the most accurate comparison I’ve heard all week. Both promise to revolutionize everything, both are about as regulated as my drinking habits, and both tend to leave people with empty wallets and broken dreams.

Want to know the really funny part? The AI education market could be worth $88.2 billion within the next decade. That’s billion with a B, folks. The kind of money that makes tech executives see dollar signs instead of students. Meanwhile, schools can’t afford new textbooks, and teachers are buying supplies with their own money.

Research is showing that students using ChatGPT are actually doing worse in some cases. German students using it found research materials easier but synthesized them worse - kind of like how I can find the bourbon aisle quickly but have trouble walking straight afterward.

Look, I’m not saying AI doesn’t have its place. Hell, I use it to autocorrect my drunk tweets. But there’s something fundamentally wrong when a company that can’t keep its chatbot from hallucinating facts wants to tell teachers how to do their jobs.

The truth is, education needs a lot of things right now: better funding, smaller class sizes, higher teacher salaries, maybe some decent air conditioning in those Arizona schools. What it doesn’t need is another tech solution looking for a problem to solve.

But hey, what do I know? I’m just a guy who’s watched enough tech “revolutions” to know that the only thing that reliably revolves is the door between tech companies and venture capital firms.

Time to pour another drink. At least bourbon doesn’t try to tell me it’s revolutionizing my life - it just makes it more bearable.

Stay cynical, Henry Chinaski

P.S. If any OpenAI execs are reading this, yes, I’d love to try your teacher training program. Just send it over with a bottle of Maker’s Mark. For educational purposes, of course.


Source: OpenAI releases a teacher’s guide to ChatGPT, but some educators are skeptical | TechCrunch

Tags: ai education ethics techpolicy disruption