Blog

NOTE: I just had a much-needed family vacation in Spain, where I also had to prepare for the biggest meeting of my career three days after my return, which I THINK went well, but I’ve always been terrible at accepting compliments at face value, so I’m not really sure! I should know for sure by the end of the month, at which point I’ll either have some amazing “Personal News” to share, or will officially start implementing Plan B for “What’s Next?”

_ONE

How a flawed idea is teaching millions of kids to be poor readers | Emily Hanford

For decades, reading instruction in American schools has been rooted in a flawed theory about how reading works, a theory that was debunked decades ago by cognitive scientists, yet remains deeply embedded in teaching practices and curriculum materials. As a result, the strategies that struggling readers use to get by — memorizing words, using context to guess words, skipping words they don’t know — are the strategies that many beginning readers are taught in school.

The “Sold a Story” podcast was the best and most enraging thing I listened to last year — which says a lot considering everything else that happened in 2025 — and I somehow completely neglected to include it (or any podcasts) in my Best of 2025 recap. This article/episode is one of the bonus episodes that’s technically a prequel to Hanford’s excellent series, and it’s a great entry point if you want to get a solid overview of the problem before diving into the whole thing.

It’s not a perfect analogy, but there’s definitely a parallel between the current “AI” push into all things and the debunked “three cueing” approach. I wonder how many AI grifters and true believers were “taught” to read that way and have no idea how wrong it was? It would certainly explain the current fascination with and growing acceptance of the “usually close enough” summaries and homogenized “good enough for LinkedIn” writing their favorite tools spit out… when prompted correctly.

We’ve seen at least two full generations sacrificed to flawed educational theories and tech-driven experiments that willfully ignore the underlying socioeconomic factors, and the beatings will continue until morale improves.

__TWO

RIP Khanmigo & Edtech Industry Dreams of AI Tutors | Dan Meyer

Khanmigo was raised with a silver spoon in its mouth, possessing advantages completely unknown to most other edtech chatbot startups. It had some of the earliest access to OpenAI’s generative AI technology. It had the backing of a hyperscalar in Microsoft and abundant cloud computing credits. It had the endorsement of national, state, and local officials. It had the phone number of some of the wealthiest people in the world. If Khanmigo died in spite of those advantages, what hope then should the rest of the edtech industry place in chatbot tutors?

I was taught phonics in school, and my now-adult kids avoided the worst of the “whole language” approach to reading, which I didn’t even fully understand at the time. I do remember being aware of, and very skeptical of, Khan Academy every time it hit my radar over the years, though.

There was plenty of research to prove that AI tutors are ineffective for educational applications long before ChatGPT was released, but EdTech’s most fundamental tenet is that teachers are the problem with modern education, and replacing them by any means necessary is the only solution. Khanmigo wasn’t even on my radar until I read Audrey Watters’ obituary last week, which pairs well with Meyer’s take if you like a little schadenfreude with your good news.

___THREE

Sal Khan’s Coming for Higher Ed | John Warner

Anyone involved in higher education, particularly public higher education or private higher ed where your institution is not insulated by wealth and privilege, should also view this project as a direct assault on their continued existence. The higher education sector and those who have historically been responsible for it (government, voters, etc. …) should pause and reflect on how what’s happened to the sector has made it potentially vulnerable to this sort of program, but we also must set recriminations aside and deal with the threat directly.

This is some dystopic Cyberpunk company school shit, and a perfect example of no such thing as failure or accountability in tech circles when there’s no limit to the money you can burn on your pet projects.

____FOUR

The Fanfare Around the Band Geese Actually Was a Psyop | John Semley

It’s an open secret in the music industry that all the numbers—play counts, followers, stats—are fake or at least obfuscated… Bots and “streaming farms” have become a marketing expense.

One of the reasons so many people think “AI” can do things as well as humans is humans have lowered the bar so much in the past 20 years, especially when it comes to digital literacy and marketing. Tech solutions to problems that were created by tech solutions is the ouroboros that will eventually kill us all.

A perfect example of how we got here is in Semley’s conclusion, where the artist who initially “broke” the story about Geese’s “fake fans” unequivocally states that she’d do the same thing if she had the opportunity.

_____FIVE

The last thing you need humans for | Melissa & Johnathan Nightingale

It matters that work be about something. That there be some actual impact out in the world that results from your efforts. It matters operationally for your business that people know what’s important and how to trade off priorities. It matters emotionally for your employees to feel like their struggles have a point and that there is value in their contributions.

The Raw Signal Group newsletter is a frequently insightful long read, and this one opens with a great story about free verse poetry before nailing an important point about the very weird moment we’re living in. The Nightingale’s appearance on Galaxy Brain is a great follow-up on the topic, too.

______BONUS