👋 friends!
This one is coming in a little later on Friday, but hopefully still offers a little light and entertaining reading as you head into the Labor Day weekend.
It has been a really big week (huge) for AI, with all of the top AI labs (OpenAI, Anthropic, Google, Meta) releasing new frontier models. It’s proving to be quite a race… (just PLEASE not one to the bottom, for all of our sakes).
In today’s note:
Parenting in the AI era: evolving education for an AI world
Connection spark: the remarkable benefits of chatting with strangers
Hands-on with AI: going (AI) analog
The whoa zone: AI toothbrushes and have we reached AGI?
🤿 Let’s dive in…
Parenting in the AI Era
Education evolution
If there’s one thing I’ve become certain of, it’s that expert opinion of how schools and education need to change to best account for an AI saturated world is still very… uncertain.
But leaders and schools are certainly trying.
You may have seen this week that NYC announced a bold new (anti?) AI policy for it’s public schools - banning student AI use for K-8, with 9-12 limited to a narrow set of approved tools coupled with 2x/year AI literacy training.
I think this will be an interesting one to watch. I like the emphasis on AI literacy training for high schoolers and a reduction of tech/screens use in class. (Fwiw, I continue to think that, at least for the short term, any testing and assessing of students should focus on in-class, device-free work).
But I worry that a blanket ban might risk a growth in cheating and inequity (I’m only suspecting, but from last week’s newsletter we saw that already over half of kids will just resort to using personal devices when AI tools are blocked).
And, I believe that the AI literacy trainings should start happening much sooner (age-appropriately of course) - from at least mid-elementary on up. And, also for teachers/staff and ideally(!) parents.

Fake, exaggerated headline but the essence 😉
MIT seeks edu solutions
MIT just released a report about AI implications on the current educational environment and found that AI is causing quite a commotion on campus:
From office hours, study groups and even in-class attendance tanking, to upending problem sets and take-home exams and even the overall grading system. Students say the rules are inconsistent and unclear. 70% also say that AI proficiency really matters for their career yet only 25% think MIT prepares them.
The report’s recommendations were some of the first I’ve read that felt somewhat ambitious, doable, and promising… you can read the full report here, but some of the recommendations include:
Start with studying and aligning around what students really need to learn in today’s AI saturated world.
Develop new assessment strategies - grades are probably no longer the answer.
Prioritize relationships and community and take a human-centered approach: shared experiences, transparency, in-person activities.
Develop tools/teams/processes for ongoing reflection, iteration, and improvement.
What I like about this is that while doable, it also isn’t just trying to put a bandaid on the problem, and it emphasizes what I think makes university so sacred: the shared learning and social experience.
But if I can opine (because, caveat, I am not an education expert)…. I don’t think we’re thinking nearly big enough yet.
AI introduces the opportunity to teach and learn in immersive and creative ways that we could only dream about. And even non-technical people can now build practically anything they think up. That opens up endless possibilities…

Setting out, as a class, to solve world hunger 🙂
CONNECTION SPARK
The people we’ll never see again
Flying out for a girls weekend this week, I ended up in two conversations with people I am never going to see again: one with the woman sitting next to me on the airplane because she had these awesome BEATs Mickey headphones (I am obsessed with Disney) and another with Taylor, the flight attendant, when I went to the back of the plane to get some water.
Neither conversation was about anything of substance, but both made the day better.
One of my favorite studies is exactly this situation. Researchers told some Chicago commuters to strike up a conversation on the train and told others to keep to themselves and enjoy the quiet. Beforehand, people predicted the conversation would be the worse ride. But they were wrong, and not even by a little.
Even better: another study showed that one week of deliberately talking to strangers measurably lowers our fear of being rejected and makes us better at conversation. Oh, and, happier and with a greater sense of belonging.
Which also brings me to our kids. We spend years teaching them don't talk to strangers, then drop them off at a dorm room and expect them to walk down the hall and make lots of friends. (The rule I think we actually mean is don't go anywhere with strangers.)
Try this: on your next errand say something to the person in line or in the elevator (what are you cooking up with all that cilantro? That’s a really neat tattoo, etc.). Even better, do this with your kids around so they can see that it’s not so scary after all.
Enjoying the newsletter? Subscribe to get a glow note every Friday.
HANDS-ON WITH AI
Make it analog
Back-to-school can make it feel like every AI study use requires putting another screen in front of our kids, but actually, it doesn’t.
Yesterday, my 5th grader brought home a two-page math study guide before a test that was happening today. The study guide was comprised of problems he had already completed and so the answers were sitting right there. Hmm… at least for my kid, staring at an already completed math work page wasn’t really working to help him get ready for the test.
So I took pictures of the pages, uploaded them to Claude, and simply asked:
“This is my kid’s study guide. Create a new printable one covering the exact same concepts at the same level, but change the problems so he has to think through them fresh.”
It gave us a clean practice round I could print out with no answers staring back at him, with confidence he was studying/practicing the exact concepts he would be tested on.
Then I asked it to “turn the same concepts into a football-themed quiz” (his passion) for an extra boost of motivation.
AI did the prep work but didn’t require my kid using a screen to study - he solved the new problems on paper, and I didn’t have to reverse-engineer 5th-grade math (embarrassing side note… I am RUSTY at basic math LOL).

THE WHOA ZONE
AI is coming for our mouths
There is now officially an AI-powered toothbrush that brushes, flosses, removes plaque, and rinses all in one go. It also has cameras to find the gaps and blast them better.
And for those of us who are more ambitious, you can even livestream the camera feed to see brushing coverage and get technique feedback.
I admit, as someone who thinks flossing is such a nuisance and dreads dentist appointments, if this actually works (and does it all in one go), I might even be willing to cough up the hefty price tag.

Source: Dyson
Buckle-up, OpenAI’s Astra model just launched
Will it bring with it astronomical chaos?
A couple of notes ago I mentioned that OpenAI had sounded its own alarm about the cyber capabilities in its newest model, Astra, and paused to shore things up. Well, it launched yesterday… apparently two weeks of work buys full confidence.
The thing that has got safety experts worried is that Astra uses, in part, something called opaque recurrence. It loops over a problem in the background, and those loops produce reasoning in a form humans can't read. Yikes…
It appears to be a big leap from any AI model publicly available - at least according to all the benchmark assessments. Right now it’s only available to a “limited set of organizations” but should be available to anyone with a paid plan “very soon.” So, stay tuned… and alert.

Source: OpenAI
And that’s a wrap!
If you've found this newsletter helpful, forward it along to a handful of fellow parents who you think might also like reading this.… we’re all in this crazy together.
And, as always, if you have any thoughts, feedback, or requests, please reply to this email with your notes - I love hearing from you!
Glow on,
Michaela
Read some other recent Glow Notes below:
