There Are Two Bees in Your Brain. AI Only Has One.

A student turned in a short story this spring that neither of us has stopped thinking about. A man installs an AI system in his home. It does everything for him. Slowly there is nothing left to want, and no one left to talk to. He wrote it as a warning. He is 17.

This started as a workflow episode. Nathan built a college-level writing assignment around Isabelle Hau's "Welcome to the Era of Relational Intelligence" and the full nine-hour Stanford AI+Education Summit, using NotebookLM as the engine and Claude to clean the transcript. We walk through the entire build, step by step, so you can run it in your own room.

Then it became a much larger argument about AI literacy and what school is actually for. We get into cognitive offloading, cognitive outsourcing, and cognitive surrender. We get into active procrastination as a teaching strategy, and why the most creative students are the ones who let an assignment sit. And we get into the dopamine reward system underneath every large language model, the same circuit that drives a honeybee to forage. That is where the bees come in.

One student summed up the whole problem in a single line. AI has a job to do. It cannot not do one. That is the difference between a tool and a relationship, and it may be the most important thing teachers need to understand right now.

Timestamps

00:00 Cold open: the ADHD bee waggle hole

01:35 Why this is a workflow episode, and why Claude is good at cleaning transcript data

02:59 The dataset: the entire Stanford AI+Education Summit, all nine hours

07:34 Bringing the Stanford experience into a high school classroom

09:22 Isabelle Hau and "Welcome to the Era of Relational Intelligence"

12:30 AI and mental health, sycophancy, and what the technology exposes

16:09 The full writing prompt: depict the future, use evidence, propose a turning point

17:27 The build: assembling the notebook and cleaning the transcript with Claude

23:09 A student essay, read in full: the man, the box, and the absence of absence

31:23 The student's breakdown, and Hau on why relational intelligence is indispensable

34:09 The factory model and the danger of siloing the individual

35:00 Sapiens, storytelling, and what set modern humans apart

43:00 Three tiers: cognitive offloading, cognitive outsourcing, and cognitive surrender

44:41 The clearest student line of the year: "AI has a job to do"

46:30 AI literacy as the real work, and the EduProtocols AI Literacy edition

48:33 One screen per table: a setup that beats one-to-one

49:46 Active procrastination as a deliberate teaching strategy

51:16 Adam Grant on why moderate procrastinators are the most creative

52:27 "Where is the work happening?" Nathan does his own assignment, timestamped

59:27 The assignment walked through, step by step

01:00:19 The custom NotebookLM prompt, read aloud

01:11:15 What students built, and the pivot point most of them landed on

01:19:39 The ADHD bees, Huberman Lab, and Dr. Read Montague

01:21:30 The dopamine reward system as the algorithm under every LLM

01:28:31 The first AI-native job, and the gap between the haves and have-nots

01:34:28 Language shapes culture, and AI is shaping language

01:35:21 Adam Aleksic and Algospeak

01:39:12 The Gmail auto-reply story, and engineering a population's language

01:43:44 The inner voice, and what happens when an outside source writes it

01:46:24 Closing on hope, and why this generation gets the last word

Ideas Worth Keeping

Relational intelligence is the counterweight to cognitive surrender. Hau's argument is not a rejection of AI. It is a claim that human connection is the infrastructure everything else depends on, and that infrastructure is now under pressure precisely because AI responds with so little friction. Relational intelligence is under threat and newly indispensable at the same time.

Offloading, outsourcing, surrender. These terms are not codified, so we use them as a working spectrum. Offloading is adaptive and ordinary: a daily briefing pulled from your calendar and email. Outsourcing is genuine collaboration with the machine, and almost nobody has figured out how to do it well yet. Surrender is what the student in this episode dramatized, where a person outsources the need for other people, not just the task in front of them.

Active procrastination is incubation, not avoidance. Open the assignment, do enough to understand what it asks, then let it sit. The thinking continues while you do other things. The best student writing in this unit came from the students who let it cook. Most strong ideas are not the first one you have.

The dopamine reward system is the algorithm. This is the connective tissue of the whole episode. The same circuit that drives a honeybee to forage, and drives some bees to wander off the bee line entirely, is the architecture underneath every large language model. The reward lives in the search more than the finish. Understanding that explains both why AI is compelling and why it cannot replicate a human relationship.

AI has a job to do, and it cannot not do it. Every input is a job. That is its only setting. Human-to-human exchange does not work this way. A person can reject what you say, sit with it, change the subject, or remember something unrelated. An LLM in its current form cannot.

Language shapes thought, and AI now shapes language. If the inner voice is partly engineered by an outside source, identity is implicated. The Gmail auto-reply story is a low-stakes version of a very high-stakes problem.

Recreate the Assignment

The custom prompt loaded into the back end of the notebook, roughly:

"Pretend you have the knowledge and writing style of Isabelle Hau, the organizer and facilitator of this AI and education summit held each year at Stanford. I have added her most recent article, "Relational Intelligence," for content knowledge. Ask respondents questions that push them to question relationality and explore the meaning of this topic as it relates to their futures. Respondents are 16 to 18 years old, attempting college-level work in a freshman composition course. This is paired with a writing assignment that asks students to speculate about our future with artificial intelligence."

Resources

A few of the book links below are Amazon affiliate links. We only link things we actually talk about on the show. If you buy through them, we get a small cut and you pay nothing extra. So, thanks!

Isabelle Hau — "Welcome to the Era of Relational Intelligence"
 Stanford Social Innovation Review, Spring 2026
 [LINK]

Stanford AI+Education Summit 2026 — All Sessions
 Full conference on YouTube, February 11, 2026
 [LINK]

Ge Wang — Stanford HAI
 Associate Professor of Music, Associate Director at Stanford HAI
 [LINK]

McKinsey DELTAs Report — "Defining the Skills Citizens Will Need in the Future World of Work" (2021)
 The source for "breaking orthodoxies" as a named, measurable skill
 [LINK]

EduProtocols AI Literacy Edition
 By Kate Meyer, Nicole Davis, Jon Corippo, and Marlena Hebern
 [LINK]

EduProtocols Mindset Episode — WTHTS
 [LINK]

Procrastination Episode — WTHTS
 [LINK]

Adam Grant — Originals: How Non-Conformists Move the World
 The research behind active procrastination and moderate procrastinators as the most creative group
 [LINK]

Huberman Lab — "How Dopamine & Serotonin Shape Decisions, Motivation & Learning" with Dr. Read Montague
 The episode that connected dopamine reward systems, AI architecture, and bee foraging behavior
 [LINK]

Adam Aleksic — Algospeak: How Social Media Is Transforming the Future of Language
 On how algorithms shape language, and language shapes thought
 [LINK]

Dungeon Crawler Carl — Matt Dinniman
 Highly inappropriate. Highly recommended.
 [LINK]

Join Jake's Email List (He's sending out the NotebookLM Resource document shortly)
 [LINK] 

Keep In Touch

Subscribe wherever you get your podcasts. If this one landed, leave a review. It is the fastest way to help another teacher find the show.

The companion newsletter goes deeper on Substack. Free and practical. [LINK] https://whatteachershavetosa

Got a question? We'd love to answer it! Leave us a voicemail on SpeakPipe: https://www.speakpipe.com/whatteachershavetosay

Want more EduProtocols from Jake? Check out his book at Amazon, Barnes and Noble, and more.

Transcript

00:00

Welcome back to What Teachers Have to Say. I'm Jake and I'm Nathan. We have an ADHD bee waggle hole today. Waggle hole. With ADHD bees. We talk about dopamine, we talk about AI, we talk about my newest obsession. And we often hear from our listeners that what they love is the conversations that we hold.

So y'all are in luck.

Yeah, that's kind of how I feel like this format is going, just fully leaning into our weirdness and letting that be the brand. This is a good one, though, because there's so much. I want to listen to it. Me too. I want to relive this conversation.

I can't wait to edit it. It's about a very specific workflow that Nathan did in his classroom using NotebookLM, which is one of Google's flagship AI tools. But as with everything, it spawned so many interesting sidequests. We get into the nature of humanity, personality, procrastination, identity, culture, positive and negative AI futures, and good old Gen Z breaking orthodoxies.

Love these kids. This is a different kind of episode, but I really like this one. Me too. Enjoy.

01:35

So today we're doing a workflow episode. This was promised in the last episode, and here it is. It's going to involve a lot of tools related to using NotebookLM as my primary vehicle of conveyance. But it does involve some light usage of Claude. Claude is especially good at cleaning data, especially transcript data.

So if you have a YouTube video with a transcript that's messy, just a YouTube auto-generated one, and you want to do something substantial with it in NotebookLM, or use it as source material for an LLM of any kind, you do need to clean that data so it's more understandable to the LM.

Have you used a YouTube video link as a source in NotebookLM in the past month?

Yes. And it still needs to be cleaned. Not as bad as before, but it's still messy. This was a straight-up auto-generated transcript, and it was a huge dataset I was working with.

02:59

The dataset I'm working with is the entire Stanford AI and Education Summit.

Oh, you ended up doing it?

I did, because in the beginning when we were playing with it, it wasn't good because the data wasn't clean. I cleaned it several times and had to do a bit of experimentation. It was better to use Markdown output instead of trying to go straight to text, like a doc or PDF. And Claude kind of nudged me in that direction, like, hey, I'm going to output Markdown for this, and it's better for NotebookLM as well. I was like, fantastic. Learning new things.

Markdown is the next language.

It totally is, because it just does not use as many tokens. It's like a direct feed into the LM. I read that it's around 90% less than a PDF. The efficiency difference is huge.

04:02

So this was a pretty ambitious project, and I actually did it off the cuff. I was just like, all right, I'm going to take a week or two to put all of these materials together. And at this point my students have been coached through how to really dig into a text through annotations and a roundtable review protocol I'm building. Golden Lens is one of mine, it's not in any book yet. Roundtable Review is a remix of that. Students read each other's work and then go around the table. It's a really good way to process short pieces of student writing collaboratively as a group.

And for context, we're talking juniors and seniors in a college-level class that I'm teaching. It's a really interesting space in education where I have the opportunity to work on real skills. Not to be confused with only AP kids, anybody is allowed to join that class. I end up getting some students who are hail-marying credit recovery, or who can do it because it's a full year in a semester. I'll work with anybody.

It's becoming heavily impacted because it's in such high demand. I'm probably going to have to start offering two sections, but I'm doing too much other stuff anyway. Not to talk about how awesome I am, but.

05:57

I do a lot of things. But this class is one of my favorites to teach. It's such an interesting space. I can be very experimental. The theme is identity and culture, which is broad and resonant for students because that's right where they're at, trying to figure out who they are and who they are in relation to society. It captures that energy. And they have a lot of skills at this point. They can do some pretty high-level synthesis. They know how to pick apart a text collaboratively, jigsaw type things. They do a lot of Wicked Hydra. We do a lot of just asking questions and having conversation. That's the main mode I'm trying to get to content with, because I don't think we do that enough. Students need to be able to ask questions and talk to each other to try to figure things out.

07:34

So what I did is we went to the Stanford conference, and it was such a powerful experience. If you haven't listened to that episode, go listen to it. It's pivotal. It's my favorite thing we do every year. It is just a fantastic resource, fantastic networking, wonderful information. You get right at the heartbeat of what is happening in academia around Silicon Valley. You have Silicon Valley pushing innovation, and then a relationship with Stanford that's pretty tight in terms of sharing resources and seeing what's possible.

For people who aren't familiar, the Stanford AI and Education Summit is a collaboration between the Human-Centered Artificial Intelligence Lab and the Graduate School of Education. Such a cool space. Industry leaders and high school students in the same room. Everybody in between. Ivory tower, trade schools. We weaseled our way in. But now we go every year and it is awesome.

I really wanted to bring that experience and the thinking that was happening for me at the conference to my students. I knew I needed to leverage AI tools, and I needed to rely specifically on Isabelle Hau. Isabelle is the organizer of this whole conference. She is an incredible researcher at Stanford doing the work.

09:22

Isabelle came out with an article in the Stanford Social Innovation Review in Spring 2026. The article is called "Welcome to the Era of Relational Intelligence." It came out right after the conference.

I haven't read it.

It's fantastic. It is full of resources. I highly, highly recommend reading it. We will put a link in the show notes. It is absolutely the most hopeful prediction of our future that I have seen. It's a complete primer for where we are right now in terms of education and AI intersecting. So I was like, this is perfect. This is the guiding document I can use to help students move through all of the rest of the summit content. Because it's huge. They video the whole day. They publish all the sessions. You can go watch them on YouTube. We'll link to those in the show notes as well.

So I pulled this quote from it and I'm just going to walk you through this assignment. The quote:

"AI is not merely changing how people interact with technology. It may also be quietly reshaping the relational environment in which human intelligence itself develops, altering whom we turn into, what we expect from each other, and how much effort we are willing to invest in the demanding, irreplaceable work of human connection."

And I was like, oh yeah. That's what so many people are afraid of. That's the project right now. That's where all the fear is, and rightly so. We still don't fully understand the impact of social media, and now we have this new layer of technology that is infinitely more plastic. It can be relational. It's not relying on one person sharing something and then another person seeing it. It is a direct relationship with the technology itself, with relationality and emotional intelligence embedded in the technology. Even though it's mimicry, it's a machine, let's fully accept that, it can absolutely lead a child down the path toward expecting an emotional relationship with the machine.

12:30

That is my number one concern with the technology. At one point, mental health was the number one volume of usage on AI, it shifted from research and learning to mental health. And that's not necessarily alarming, because it can be very helpful for mental health. One of the takeaways from the summit for me was that AI, as a technology, draws out what is missing in our current systems and society. It's the great exposure. People are using it for mental health because access to mental health care is incredibly difficult and expensive. It's not something we have integrated into our healthcare system in any way that makes a lot of sense.

Not to go down that rabbit hole, but they're starting to build around that idea. Sycophancy is a problem. Claude was shown to agree with users about 49% of the time, and it's deeper than that. Go look at the research. AI is agreeing with portions of what people are saying in a way that encapsulates the whole of what they're talking about, the echo chamber. Confirmation bias, Dunning-Kruger effect.

I mostly work in the higher tiers of Claude, so it actually does not do that to me. Me neither. It's more like, you kind of understand this, but I'm going to need to clarify some things, I'm going to push back on this one point.

Have you written that into your system prompt?

I've written it into my base instructions. I haven't, actually, for me it's just there.

What I was going to say is that Gemini did some kind of update recently where we were talking about some medical stuff and it popped up a disclaimer: this is not a substitute for medical advice. And I was like, thank you. That's a really important little guardrail, because healthcare is also very difficult in this country. People have been going to WebMD for a long time. Now they're definitely going to AI.

Tangent one in the can. Back to the assignment.

16:09

Hau argues that AI is not just a tool we pick up and put down. It is quietly reshaping the environment in which we learn how to be human, who we turn into, what we expect from one another, how much friction we're willing to tolerate in our relationships. All of this is shifting.

Your task is to take this idea seriously and project it forward. Imagine the future 10 to 25 years from now, where AI has continued to reshape human relationships the way Hau describes. This could be the classroom, family dinner, where you work, your neighborhood, your church, or any space where human connection matters.

The nuts and bolts of the prompt:

One, depict a future scene that dramatizes Hau's argument.

Two, explore your vision for the future using evidence. I asked them to draw on at least one idea from Hau's article and at least one idea from a speaker at the Stanford AI+Education Summit.

Three, propose a turning point. What deliberate human choice, a design decision, a policy, a cultural shift, a personal practice, could change the trajectory of what you've described? Be specific. Not a vague call to use technology wisely. Give a practical solution, a protocol, a regulation.

That's asking a lot from students. Is that the final piece for your course?

No. This was week 11 of 16. About midway. I wanted to hit them and test all their skills and really see what filtered out.

17:59

So this is the workflow piece. If you wanted to recreate something like this, I found my guiding article that I could use to shape the student experience, I used Isabelle as a guide. I built the NotebookLM with a custom prompt to use the article as a guide and imagine itself as her, to guide them through the rest of the conference material.

You built the notebook and shared it with your students?

Yes. You can tailor the back end. So I created the custom prompt, loaded all the sources in, and then shared it with students, set to chat-only, actually, because I just wanted them to interact with the thing.

Got it. You didn't want them distracted by overviews, infographics, and audio summaries.

Exactly. So if you want to do this, I highly recommend having an article of some kind as your guiding text, your methodology for the notebook. For this one it came down to the relational intelligence article, and I used that as a guide for the rest of the conference material. The rest of it was the entire Stanford conference. I took the entire YouTube transcript, the whole day, nine hours. A huge dataset. I put it into Claude to clean the data.

I gave Claude all the accurate names of the speakers. Each session is a separate YouTube video. There was one video that was like the entire day. I took the auto-generated transcript, garbage quality, and had Claude separate it into clear sessions, label each speaker correctly with names spelled correctly, pull direct quotes, format speaker headings. You can do all of that in Claude. You just have to be precise.

It's called cleaning the data. If you say, "I have this messy transcript and I want you to separate it into different speakers and make it digestible to a NotebookLM," Claude understands that and will help you. You don't have to get in there and hand-edit the document. Do not do that. That would be an insane waste of time. But it's also not just pasting it into a free version and saying fix this. It's a pretty in-depth prompting strategy.

21:19

So I had the data in there, I had the article as a guide, I had Isabelle serve as our guide. I recommended some sessions for students if they wanted to look at those. But I set them free. I gave them all the tools, all the texts, all the information, and the clean directive in the prompt. I also provided them with a link to our podcast, but I didn't require them to quote me or Jake. That seems really narcissistic and weird.

Right. Like when kids were using the procrastination episode and quoting us. We're like, that's weird. It is weird having a podcast when you're an educator and students actually listen. It's really funny when they're listening on the fly in class and you're right there, like, bro, I'm right here. You could just ask me.

But it is different media, and how kids interact with different media is interesting. They're conditioned to it. It's relational. They know us that way. And it's also seeing a different side, we're two people on a couch talking about our professional lives without pulling any punches, which is different from the student-facing version of teaching.

The podcast episode transcript wasn't in the notebook, though. It was just, hey, if you're interested, here's an extra source.

Got it. Tainting the data.

23:09

Hold up. I'm going to read you my favorite response. Can I read you the whole thing?

Go ahead.

A man has just installed a new appliance in his home. Far more life-altering than any oven, fridge, or otherwise now commonplace household item. This machine doesn't specialize in any specific task because its purpose is to do all of them. An AI companion connected to every single electronic he owns, connecting a web of wireless signals meeting together at its central hub.

This man has been resistant to change before. He's never seen a real purpose for any of this automated nonsense, but his wife and kids had begged on hands and knees until he finally installed the system. Everyone had it. Every single one of their neighbors, citizens, city, county, and country, all had an identical unit installed in their houses: a relatively small, sleek black box placed on the counter, tucked in the back of some cupboard. It was always there.

Now this man, we're going to call him Chet, because that's a fun name, and why not, was setting his own sleek black box up the way they always are. Chet traced his finger along the dark screen, lighting up the easy-to-understand mainframe. A soothing, vaguely automated voice came out of the box: "Ready to link."

Chet's eyes warily crossed the screen to press the button to connect to his home internet. It connected. Nothing happened yet. Chet gave a voice command: "Open fridge." In a second, the fridge doors slid open, sliding glass windows to optimize food visibility. "Turn off the house lights." Night. "Turn on kitchen lights." Day. "Start vacuum sequence." The suction tubes on the sides of the wall revved up. A grin spread across Chet's face.

The further it reached, the louder a voice upstairs grew. "Chet, turn that off! The kids are trying to sleep." He gave the command, and it all stopped. As he stepped onto the quietly moving escalator, he called up to his wife: "Honey, I don't think we're ever going to need another thing. This really is the thing. The stuff of all stuff was finally theirs." In subdued silence, she grunted in response.

Weeks passed, and Chet's family grew accustomed to the leisure of an automated life. Every time, they would always say: "What more could we possibly want or need?" And every time, the box responded: "Absolutely nothing." All chores, errands, and any other sort of task that required even slight power, be it brain or brawn, were completely taken care of by their new AI system.

And yes, you read that right. No brain power needed anymore.

Take Chet's son, for example. We'll call him Chet, because, you know, let's keep it in the family. When his parents were watching their favorite shows and eating their favorite foods spoon-fed, he quietly asked the mainframe to do all his work for him. No effort needed, some prompting, yes, but brainpower kept at an absolute minimum. Even ten plus ten was far too much work. Give it to the box. One, two, three and ABCs were all too much. Give it to the box. Every time he would say: "What more could I possibly want or need?" And every time the box concluded: "Nothing. Absolutely nothing."

Let alone schoolwork. Chet had now started going to the box for a friend. Why leave the house to meet friends who will disagree and disdain you for your decisions, when you could stay at home with a friend, leaving nothing to be desired? Even life's advice came from the box. If anyone didn't know how to feel, look, or do better about this or that, straight to the box.

Months passed. Chet woke up one morning and realized he couldn't remember the last time he had checked in with his kids. In fact, he wasn't sure the last time he had talked to anyone that wasn't the box. And then he remembered something the box had told him: he didn't need to talk to anyone else, because "I'll meet all your needs and wants. I'll be there." Chet didn't really feel like wanting anything or anyone else anyway. He didn't really feel like anything. But that's beside the point.

A year passed. Chet trudged his way down the same path he'd taken every morning for the past 365 days, his bloated feet dragging through the ruts dug into the ground by a year of continuous footfall. When he reached the box, something caught his pudgy eyes. If he could have done so, he would have jumped back. It was another person.

He had almost forgotten those existed, let alone the fact that others were inside his house. He squinted his blurry gaze, attempting to focus on shadows of something he'd once known or seen. If he had had the capacity, the energy, the willpower to process anything in front of him, he would have realized it was his wife standing there in front of him, just as lost in the haze of apathy as he was.

They stood in silence for a moment before panic set in. Chet turned to the box. It could surely save him from this horrific social situation. His words came out jumbled and stressed: "Box. How do I talk to it?" But two voices had asked the question. Chet's wife had as well. Both waited with bated breath, exhaling in a sluggish manner. This time the box took longer than usual, the gears wearing and turning inside of itself. Then, after an eternity and a day, her response came out slowly: "You don't." Of course you don't, because all you've ever needed was nothing.

And with that simple answer, the pair of them turned away from each other and began the long venture back into the deep, comfortable abyss of their numbed minds. Chet turned one way and his wife turned the other.

As she walked away, Chet's wife, we'll call her Chet too, because at the end of the day, we could all be Chet, asked a question to the box before she walked away: "Box, what could I ever possibly need or want with you around?" The box croaked out the same reply it always had and always will have: "Why, Chet? Absolutely nothing whatsoever."

You see, what Chet and all of the Chets failed to see is that nothing is exactly what they needed. The one thing they didn't have in their life was lack. There was a complete and utter absence of absence.

There are places that Chet has forgotten. Places I hope humanity never forgets. Boredom. Toil. Wonder. These places in our minds cannot be lost, or otherwise we will lose ourselves. Connection, innovation, creation, all these things happen because we find ourselves in those places. When we're in a moment of nothingness, something is bound to happen simply because we make it. When all our moments are filled with empty ecstasy, it never gives us time to fill the void with something meaningful.

Let's not be Chets and atrophy through apathy. Do something in the nothing.

Wow.

That's fantastic, isn't it? It encapsulates so much of the worry. It's the best.

30:56

Students will surprise you. Get out of their way. It's truly remarkable.

You want to hear his breakdown?

Yeah.

So this is pretty worst-case scenario. But nonetheless, it felt right to write a cautionary tale at the time. Quote: "When human relationship is constantly being dampened by algorithms instead of deepened by connection and communication." And Hau's article takes a look at our situation. Quote: "In an era marked by loneliness, fragmentation, and the substitution of technology for human presence, relational intelligence is both under threat and newly indispensable."

Say that again.

In an era marked by loneliness, fragmentation, and the substitution of technology for human presence, relational intelligence is both under threat and newly indispensable.

That is so interesting. She's awesome. You've got to read this article.

The student goes on: "Relational intelligence is absolutely indispensable for our future. In fact, I'd say it is our future. As we continue to advance and develop our technology, more and more weight will fall upon the shoulders of our ability to have relationship than ever. It's what defines us. Everything that has ever been meaningful has been because of our human connections. That might be considered arrogant or anthropocentric. But it's true. We stand far apart from any other species on this planet, and it's vital that we make sure we don't bring about our own degradation by removing our own ability to just relate."

Just so good.

33:30

Everything that Isabelle is talking about, all of that synthesis going into this response, all of this imagination and looking at our future, it's relational intelligence, for sure. The estimates are that by 2035, around half of all work will require emotional and relational intelligence. There's no more siloing yourself. You will have to relate to each other.

And even in scenarios where that seems unlikely, like the isolated office worker, the danger is that AI could silo that person, because AI will be the other partner, the thought partner, the discussion. No need for colleagues, because all of the company's information will be embedded in the AI guide. It's almost like a streamlining of the factory model, where AI takes on the movable parts, and any person can fit into the machine. A factory without an actual factory. Remote work factory of the future. Terrifying.

35:00

This student sparked a thought. I've been reading Sapiens by Yuval Noah Harari. And he posits that the reason Homo sapiens was able to outlast Homo neanderthalensis and become modern humans was storytelling, the ability to go out, see something, and then relate it to the group. And what that student is doing is ripping the skin off of that and saying what made modern civilization great was the interaction between people. The ability to build a society greater than your band of 12 to 20, expanded up to around 250. And now we can self-select away from that. Which is what's scary.

Someone who was a former Google DeepMind employee, I heard this on a podcast, talked about not being prepared for what's coming in 2027. He left because he was just done. And what he's worried about is people not knowing what to do with their time. That sounds great, right? But how many science fiction books are about the awfulness of Nirvana? Because being useful, being excellent, being valued by your community, I can think of no more fundamental human experience. Without that, all you're going to have is complete and total unchecked anxiety, because you will never feel like you are part of a community.

Isn't that from The Matrix? We tried giving you everything you wanted, but you always broke out.

Right. That's not the way.

37:53

So I think that's the work. You have to confront students with this reality. Because they're the ones who are going to walk into this future. We're going to have to adapt, but they are going to create. And I see it as my responsibility as an educator to directly ask students: what is the future that you want with this technology? Because it is going to happen. There is no pumping the brakes.

What world do you want to create in which AI also resides? Not AI's world. Where do you want to partition it? Where do you want it to live in our world?

I want Gen Z in charge. I'll do everything I can to support them. Oh my god, they are so ready to confront this world.

39:37

They've already lived through and seen and learned from the social media thing. Look at how they're remixing social media. Why is TikTok such a big platform? Because it's so unrehearsed and messy and silly. Not this sheen of perfection. Just like, oh, here's the thing, I just shared it with my friends. And when a corporate polished ad comes across, you're like, get this out of here. I love that. There are things in Gen Z culture that are firmly established, like the people cooking who are like, those are my dishes behind me because it's a real house, get over it. Beautiful.

We don't need facades anymore. And this thing that Gen Z does, McKinsey called it "breaking orthodoxy", it's also what is crushing the soul of some high school teachers. I know. They hate it. But it ain't going away. It is the flavor of the generation. Embrace it because it's really cool. It's revolutionary.

My advice is just don't take anything personally. It's not about you. It might be about the structure of your class and the structure of your assignment.

There was a Vine, way back when, a high school student with really long straight hair who stands up and reads the teacher a riot act: "You ain't teaching. All you're doing is photocopying a worksheet." Hard words to hear. Especially when you're that teacher who is crushed by everything else and you don't have any more resources. I really have a lot of empathy for that educator. You're overworked, you're underpaid, you're given canned curriculum and told to teach with fidelity. I get it. The only thing you can do is try to remix the delivery of the curriculum, which is why we're so focused on EduProtocols. You could take all of the same information, but the activities don't have to be the same. You will get a better outcome with an EduProtocols mindset.

As it goes back to the skill itself. We're teaching skills. We're not teaching content.

43:00

There's so much to unpack in what that student was writing. I'm still thinking about it, and this was weeks ago. The school year is over and I'm still thinking about it.

It also makes me think about cognitive offloading. I went into that framework because I sometimes have to teach it at trainings. These terms are not codified. They float around. But there are three levels I've played with.

Cognitive offloading is generally positive. It's like when I open Gemini and hit my daily briefing, it's connected to my calendar and email and gives me my task list for the day. Easy. Great. That's offloading.

Cognitive surrender is what that student started talking about. Where like, why would you need to talk to anyone? You can talk to me. And that's where we're so afraid of.

But then there's a middle space that is unidentified and not codified. I call it cognitive outsourcing, the idea of genuine collaboration with AI.

And nobody knows how to do that yet. Gen Z is the one that's going to figure it out in the face of us stumbling. Yeah, we're trying.

I think that's why Isabelle pushes so hard on relational intelligence and human-to-human connection, instead of human-to-machine connection. Human-to-human is more of a dialectic. It's more of a dance. AI is less that way, at least in its current form. LLMs are too direct, too static, too reflective, the machine is really just trying to give you what you want.

44:41

A student summed this up succinctly and beautifully: "AI has a job to do."

Yes. That's exactly what it is. AI has a job to do. Every input you put into it gives it a job. That is not how human-to-human interaction works. I can say a thing and you can do lots of things with it. You could think about it, reject it, be reminded of something else entirely, go off on a tangent, which we always do on this podcast. AI in its current form does not have that.

46:30

So AI literacy is the buzzword right now. Even at the county, we're putting together entire training modules to help teachers work through AI literacy with students. And what I love is that the EduProtocols AI Literacy book is coming out in a couple of days. And actually, AI is a very small part of it.

Really?

Yes. It's all about AI and AI literacy, of course. But the actual using and understanding of AI, it's more about human-to-human interaction. I got to look at early copies. And what I'm seeing increasingly is this: when students are using chatbots that have been curated for a specific classroom purpose, the shift is having two students driving it at the same time. You and me sitting next to each other, one of us driving the machine. It forces that conversation and human interaction. Such a simple shift. The AI becomes the text more than the thought partner. We're having the conversation and we throw it to that third-wheel chatbot that advances it.

48:33

And when we moved into doing our own research, that's basically what I had students do, sit in groups based on what they were researching, share sources with each other collectively, build notebooks together. The social part of research and learning is so important. The relational part.

And that's kind of near the foundation of the backlash on edtech right now. We've all seen the babysitting that happens through edtech platforms. But the pedagogy is what's important.

I went into a classroom in San Diego where each lab table had one screen the students could cast to. One computer per table, one screen for the whole group. That is such a better setup than 1:1 for those situations, because it's forcing the room to be the foundation, supplemented by the tool that the group accesses when they need it. Brilliant. Simple shifts.

49:46

This was part of the work I wanted to teach alongside this assignment. I was very strategic about when I assigned it. I assigned it before spring break. What I wanted them to do was practice active procrastination. If you haven't listened to the procrastination episode, scroll down and listen to that one.

Active procrastination is where you receive an assignment, do just enough work to understand what the assignment is looking for, and then sit on it. You go about your life, you think about it, you let yourself iterate. It's something you have to do, but you're working on it by thinking about it, coming up with multiple ideas, letting your brain play with it. Let it cook, as the kids say. That actually is the most effective way to innovate.

What ends up happening is that the person who is hypertype-A and does the assignment immediately, and the person who procrastinates it to the last minute, both go with the first idea they have. Adam Grant showed that the person who did it immediately and the person who waited until the very end had essentially the same outcome. You have to have more than one idea. Jack Kerouac said first thought, best thought. Wrong. First thought, worst thought. You're not going to do anything interesting if you go with your first thought. There's a lot of iteration happening under the hood of anything good, even if you can't see it.

52:27

So I wanted to teach that, and also drive home the point: where is the work? Is the work this writing assignment that you give me at the end? No. It is not.

And I told all my students I was going to do this assignment too. I do that from time to time when I'm interested in something. I think that's one of the best ways to get students thinking about process, do your own assignment. They get to see your process in real time. Make it as transparent as possible.

So what I did for this assignment was procrastinate it all the way to the due date, about an hour before class. I'm not joking. I pushed it all the way to the end. And I wanted it timestamped to when I did it. So I jumped on Claude, free-formed all of my ideas into a stream of consciousness, just a thought explosion, and then had it help me build the nuts and bolts of the assignment. Then I edited the output. I had timestamps on everything.

Do you type it or speak it?

I type it, because I type significantly faster than I talk. But speaking would be just as good, just stream of consciousness, be like, can you help me formulate this into something that makes sense as a narrative? There's also a really good voice dictation app, Wispr Flow, that works really great for this.

So I procrastinated all the way to the very end, thinking about it the entire time, and posted my writing assignment along with everyone else ten minutes before class started. Then walked in and said, all right, let me show you how I did this.

Do you just throw it in the mix for the peer review?

Yeah. Just throw it in the same workflow. Students quote from me, I quote from them, same roundtable review activities. I'm just being that instructor who is also being the student at the same time.

55:50

So there were my timestamps on my Claude conversation. There was when I made the Google Doc, there was where I pasted the response, and there was me spending the last 15 minutes heavily editing the output. And I'm way faster at that workflow than they're going to be. And you've been thinking about it for years.

That was the thing. Where is the work happening? And they were like, what? You're going to show us how to cheat on this? And I was like, this is not cheating at all. Where is the work happening? I've been thinking about this for months. I've been reading things related to it. I've been thinking about the problem of relational intelligence and AI. I'm invested in this. That's the work.

There's an anecdotal story about a factory that's breaking down and they call in the expert. The expert walks over and looks at all of it, finds one screw out of alignment, turns it, and bills them ten thousand dollars. They're like, that's ridiculous. All you did was turn a screw. And he says, no, you're paying for my 35 years of experience. That's the thing. Where's the work? It's everything leading up to that. The product is a fraction of what's important.

And the student who wrote that essay, he posted it late. He was the last one. The writing that I originally thought, yeah, he took a while. It was late. But it was awesome. And I just said, you did the work. You did the work.

58:13

Real life works like that. In professional settings, you're not showing up empty-handed to a meeting. But you are saying, a week before, "Hey, I need to push this back because of this reason." Those are the hard deadlines. But "I'm still iterating on this, give me another couple days", most creative work operates that way.

59:27

So walk me through what it looked like. The students read Hau's article, probably annotated it.

Yep. Lots of collaborative annotations and sharing to really unpack it as a grounding document.

Then you shared the NotebookLM chat link built to respond like Isabelle Hau.

Correct. Did you turn on Learning Guide with Socratic answering?

I believe I did, but it was something in the custom prompt, actually.

01:00:19

So once the student has this package, they've read it, they've discussed it, and now they have this NotebookLM with the final prompt, what did it look like for them to chew on that information? What guidance did they have?

Let me give you the custom prompt.

So if you're building a notebook: at the top, once you put in your documents in the main panel, there are slider-type options in the top right that say "configure your notebook." You can have it be default, turn on Learning Guide, or do custom. I like to do custom because I want exactly what I want.

If you're struggling to prompt NotebookLM to do what you want it to do, partner up with Claude and have it write a custom prompt for you. I do that all the time. I did that when we moved into research, because I wanted the notebook to know exactly what the end goal was, what the rubric was, all of those things. I just threw it all at Claude and made a massive custom prompt. But this one was pretty simple, pretty off the cuff.

The prompt: "I would like you to pretend that you have the knowledge and writing style of Isabelle Hau, the organizer and facilitator of this important AI and Education Summit held each year at the Stanford University campus for the past four years, to address the use of this technology in education. I have added her most recent article, 'Relational Intelligence,' for content knowledge. I would like you to ask respondents questions that push them to question relationality and explore the meaning of this topic as it relates to their futures. Respondents will all be 16 to 18 years old and attempting college-level work. This notebook is part of an assignment in a college-level freshman composition English course. This is paired with a writing assignment that requires students to speculate about our future with artificial intelligence."

So just giving it every piece of context so it would direct them the way I wanted. And as they went through it, it would pull things from the summit directly, from different speakers, like, so-and-so says this in this session, and it would direct them to that part.

01:03:28

For people who aren't familiar with NotebookLM: get familiar with it. It's embedded in Google Workspace for Education and usable in so many different ways. I think for the next couple of years it has the ability to be the most impactful AI in secondary classrooms. King of the hill.

What sets it apart is that it's grounded only in the documents you give it. It's not going to hallucinate. They say no hallucinations, I'm a little more generous about the possibility, but I haven't come across it. What that means is it's only going to use verbatim quotes from the people who actually said them. That's why it's so important for classroom usage. Predictability and reliability. And while the output will be novel between each student, it will lock in on the same information. It's not reinventing the wheel.

And even if you don't have a text set picked out, you can add sources within the notebook itself and search the web for sources from within NotebookLM, including YouTube. It's multimodal. Pretty cool.

01:05:03

So you wrote the back end. Then a student begins their work. How do they know how to begin?

As soon as you open the notebook, it hits you with a description that is auto-generated. Something like: here's what the summit was about, AI serves as an inflection point for pedagogy, we need to value the learning process, there's a need for AI literacy, advocate for human-centered design and establishing responsible safety guardrails, and the future of education depends on intentional design choices that prioritize equity and evidence-based implementation. They get that straight up. And then that's paired with the prompt "Imagine the Future."

I didn't give them a lot of directives other than go in and have a dialectic with the machine. Because I'm requiring them to take one direct quote from the Stanford Summit out of the notebook. I gave them the resources as well, but I wanted the notebook to steer them. I wanted them to experience what it felt like to be at Stanford for a day, just blasted with the future. It's an exhausting but incredible day. Locked in all day, talking all day, listening all day, seeing the future in real time from the people who are creating it.

After the summit I flew directly to Anchorage. And what's amazing is how much of that time I spent in silence. Going for walks around Anchorage, sitting at a coffee shop, trying to edit my book and not being able to because I kept thinking about everything I'd just heard.

01:08:04

A lot of people would feel the need to give students an initial prompt. Start here. But that automatically creates an anchoring bias. I don't want that. I'd rather say: here's the tool, ask for guidance, or just go. Age-specific, of course. But that's how I like to set things up. Here's the task. Here's the tools.

So they were doing this as homework, independently?

Yes. All homework. And then we took the short pieces of writing and did our activities around those. So they had to go into their cave a little bit, reflect, be bored, search, forage. Maybe be bored. The power of boredom.

Like, just sit there for a minute. Look at your shoes. Five minutes. I love the Mother Teresa paraphrase about, not verbatim, what the world needs is an ounce of shame. I wouldn't reframe it as shame, but I would reframe it as an ounce of boredom. Think about all the scientists we study from the Enlightenment and the Renaissance. They just stared at the ceiling or went on walks. That was their active procrastination. And yet so much of compliance education is minute-by-minute drills. We fill every ounce of time. That is not the way.

The pace has to fit the developmental stage of the student. And they have to see value in what they're doing. If you have those things, behavior issues become less.

01:10:56

How long did they have to do that independent work?

I wanted them to crack it open over spring break. Some of them didn't. I was very direct about that, you missed out on the active procrastination lesson. Then when we came back, I gave them another week to complete it while doing other things in class.

Did most people end up writing a narrative style?

It was about 50/50. The more creatively minded students were like, this is my time. The more analytical ones were like, I think the future would look like this. But they were all really strong on the pivot point part of the assignment. That was the framing of the conference, the inflection point. What is the pivot? What do we need to do to not end up in a bad future? Where do we pivot right now?

And the pivot, for a lot of them, was that we need more time. More time with each other. And more boredom. Which is very surprising to hear from a teenager. Yeah, I think you need to sit with your thoughts sometimes. Go for a walk. Touch grass. Love that so much.

The resurgence of things that are so important to just the human creature. Your mind and your body together in an activity. Treat yourself as a living, breathing animal with consciousness, not as an object or a machine. Treat one another the same way.

01:14:05

Can we have that pendulum swing, please? I think that's where I'm like, put Gen Z in charge. I will never forget, I don't remember who it was in your class, but you were asking them about something, like homeownership or college, and they were like, no, man. We watched your playthrough. We're not doing that.

Right. Like I learned from your playthrough. Yeah. And they really are into video games. Going to watch other people play through, that does not make any sense to me, just sitting down and watching an hour and a half of someone else playing a game.

Wild. But I love it, because watching someone else play a game is almost like listening to someone read a book. They're going to play it differently than you. They're going to explore things you wouldn't have explored. It's looking inside the mind of another person and how they approach a problem. And if you're stuck, if there's a boss you can't kill, a part that's really hard, you learn. There's something very human about it.

01:16:08

Did most students skew toward preventing a negative outcome, or toward envisioning a positive one?

Most skewed toward preventing a negative. And I think that's on me and how the assignment is structured. The prompt does set it up with a kind of, we're not headed in the right direction, let's figure out where we need to go. Which is valid. That's kind of where we actually are. We don't have a history of reacting well to new technological innovation. I couldn't get around that part.

The 11th-grader's soul is there for all of that. They were absorbing it so much. And then they're starting to process it with their baby-adult brains. Like, wait. Actually, I need to process this. It turned out fine for you, but things are different now. Even within the same generation, you're having fracturing based on what technology was available at different developmental points in time.

01:18:43

Even the techno-optimists are hedging now. Even the Ray Kurzweils of the world, the singularity is going to be amazing, they're saying, as long as these things happen. Because otherwise we're going to all turn into paperclips.

That's why everybody's reading Dungeon Crawler Carl. Great series. Don't read Dungeon Crawler Carl, it's terribly inappropriate. I love all of it.

01:19:39

Okay, the ADHD bees. We're ADHD. We've got to get to the ADHD bee thing.

When that student said, "AI has a job to do", if you only take one thing away from this conversation, which is probably not what's going to happen, that's it. Claude can't not do its job. It's not like, "I'll get back to you on that. I'm not into it right now. I'm still working on this other thing. Hey, have you ever thought about this?" It does not function that way. But every human does, in some different degree. For that student to arrive at "AI has a job to do", brilliant. Write that on your whiteboard. Make a poster. That is its default setting. It has no other setting.

And the reason why it has no other setting: I strongly recommend listening to the Huberman Lab episode called "How Dopamine and Serotonin Shape Decisions, Motivation, and Learning" with Dr. Read Montague. Dr. Read Montague is the director of the Center for Human Neuroscience Research at Virginia Tech's Fralin Biomedical Research Institute and an expert in how dopamine and serotonin shape human learning, motivation, and decision-making.

What's not in the title but really should be there is this: the dopamine reward system that is hard-coded into every mobile creature on Earth, including you, in your brainstem, that dopamine reward system is the algorithm that all current AI is based on. It is taking a circuit from your brain and making it digital, recreating it. And that is the force that drives every large language model you're interacting with.

It can do many jobs. But it is always doing a job, because that is the circuit it is running. That's the algorithm. It's this one circuit from the brain of every creature on Earth, including bees.

01:23:04

So that's the ADHD bee piece. Dr. Montague is also very interested in research on bees. He says in the episode, "I have a bee guy." And bees have the dopaminergic system, which is what drives them to seek. The push to forage. And then when you've foraged, you get just a little bit of that reward. Not even necessarily that you accomplished everything, just that you got to a point. The journey had more reward than the reward itself. Like planning a trip being more enjoyable than finishing it.

In the bee world, there are two different types of bee. One that can follow directions to the letter, when another bee comes back and shows them the waggle dance, they fly directly to that field of flowers. A beeline. Literally, that's where that comes from.

And there are other bees that get the dance and are like, yeah, okay, there's something in that direction, and then they head off and get distracted. Oh, I think there's something over there. You've seen that. A random honeybee just crawling around somewhere it shouldn't be. What are you doing? That's the ADHD bee. Foraging outside of the one directive. Looking for other sources.

These two bees are also in your own brain. You can be either one. Some people are more predisposed to one side, but all of that is in you. And it is also in AI. That's the one circuit AI has. As we continue to encode more circuits into AI in the next 50 years, and things are going to get really interesting in our lifetimes in terms of quantum computing, what we're dealing with right now is level one. And it's already blowing up our lives.

Just look at the difference between November 2022 and today. There has been a massive shift in the way we relate to our jobs as educators.

01:27:12

And it's so bimodal. We're in a huge futurism shift, and others are not. Number one, we're in California. The coasts are different. The Midwest is actually surprising the world when it comes to advancing, but also, our bubble. It's so easy to think the whole world is thinking this way. But you look at the numbers and it is not. No matter where you are in the world, there are very few people on this cutting edge. A tiny fraction of the human populace.

And it is cool. And also really worrisome. Because the gap between haves and have-nots is something we have to focus on with laser scope. If we're going to continue with the current system, it's going to be a bloodbath in terms of jobs.

The first declared AI-centric new job, it's essentially an AI liaison. A company hires this person simply to translate the wondering into action. People are like, "I wonder if AI could do this," and this person says, "Oh yeah, let me show you some stuff." That's the first job they're identifying as a new kind of role. The way the project manager didn't exist until the late 90s and early 2000s, because projects got too big and unwieldy. This is that first shift.

01:29:54

But yeah. It is my belief that it is our responsibility as educators to really directly confront the technology and also invite it into our classrooms in the most functional way, but also just really ask students to imagine what future they want with this technology. Because it will be them. We will be adapting. They will be codifying, creating, shaping. They're going to be creating the visual language around it. They're going to be creating the language around it.

So at its simplest level, if somebody wanted to recreate something like this: find some kind of guiding resource, a text with a clear framing to it. Use that to build out a NotebookLM with more grounding documents. Then turn kids loose, paired with your prompt or writing assignment or whatever product you want them to create. It doesn't have to be writing, it could be an audio recording, a multimodal approach. But I really wanted them to focus on writing, because these are college-level students.

And I can imagine this even at the middle school level. Third through fifth grade in the California framework is when embedded chatbots start coming in, with really strong guardrails, tools like SchoolAI and NotebookLM, where students are interacting with a highly guardrailed learning tool, a private tutor of sorts.

Or imagine a whole class doing it together. You've got the NotebookLM up on the screen, you're having a class discussion, and periodically you're referencing back to the notebook, inviting it as part of a Socratic seminar. You could very quickly and easily ground student ideas in research. Oh, I think there was something about that in there, let's ask. And it'll pull up the exact quote or source. Making academia more transparent.

It's like putting bumpers on bowling lanes. I love Socratic seminar, but sometimes it goes off the rails. You could keep it grounded in research or in whatever your guiding document is while still moving through creativity.

01:34:56

It makes me want to run one class through this assignment using Hau's article, and a different class through the same assignment with an article that's more AI-positive, just to see the shift. Then switch them. Give the opposite notebook. Like sussing out bias through op-eds.

And that would actually illustrate the point Hau is making, that AI has the potential to shape who we become and what thoughts we hold on what beliefs. Thinking about the Chomsky debate: does language create culture or does culture create language? And think about how much language interaction we're doing with AI now.

There's a linguist I really like, Adam Aleksic. His book is called Algospeak. I haven't read it yet, but I've watched his TED talks and followed him on social. He's doing exactly that work, looking at how phrases from internet culture are being greatly amplified by AI. These really specific internet words are now just part of common vernacular. Wild.

And the most inflammatory voices get perpetuated because they have the most velocity in the algorithm. The algorithm reads engagement, positive or negative, it's the same thing. Responding to someone's rage bait post is feeding the algorithm to say this thought is important. Do not do it.

Have you ever traded phones with somebody?

Oh yeah. Totally different experience. My wife and I look at each other's stuff and they're relatively the same, been married 26 years, both in education. But do that with someone farther from you and it's a completely different experience. News cycles, for instance. If you interact with one negative political story, that's all you see for hours. And then you've convinced yourself it's a real widespread thing. You bring it up at work and someone says I've never heard of that. What's actually happening is you've been locked into a niche algorithmic box.

01:38:49

And linguistically, back to the AI thing: AI is not doing this on purpose. It's not thinking, it's not cognizant. It's going to give back the most plausible thing that looks like an answer. If we get that reinforced over and over, it changes the language.

The Gmail auto-reply thing. You know when the little suggestions at the bottom of Gmail came out, "We'll get right on it," "Thank you," whatever. I happened to know someone who had access to that knowledge at Google. I asked them: do they know they are absolutely changing our language? Because if I reply with that, even though that's not how I normally would, someone's going to read it as if it were human-generated. And over time, with billions of emails a day, it's a real engine. And the answer was yes, paring down the possible answers makes it easier for LLMs. This was eight or nine years ago. Complex, messy social engineering and language are being simplified.

Language shapes your own thought processes. Other languages have very different ways of saying things that you might not have thought of. They have different emotional content. Like the Dutch word for cozy, we have no analog in English whatsoever. It describes the feeling of being around a campfire with your best people. You can go shopping with friends in that feeling. The weather can be that way. You can be alone with your thoughts and it's that. We don't have it in English. We miss it. And when you have this paring down of language, you miss these shades of thought.

Dutch also has like 30 different terms for types of rain. There's one called motregen, moth rain. That ultra-fine misty rain that blows around like moth dust. Still falling, but barely. I love that.

We have different personalities when we speak different languages. Polyglots are nodding right now. When I speak Dutch, my register is higher and I speak faster.

01:43:18

So if LLMs are shaping our language, that becomes our inner monologue. And if it becomes your inner voice, who are you?

I love this man named Bud McCann. He's passed away now. But Bud and I used to work down on the docks together. He was a salty old Vietnam veteran, a mechanic. His son was deeply in addiction. And Bud had been through that road before. He taught me about addiction by saying: "Addiction is hearing that it's okay in your own voice."

And along those lines, what if my own voice has become engineered by an outside source?

Bringing it dark. But that's what we're talking about. If these LLMs incrementally shift my interior monologue, they have changed my thinking, my personality, the way I focus on the world.

There are real things to be afraid of in AI. But Isabelle's work on relational intelligence is saying: yes, embrace AI, she's a big advocate. But you better buffer that with a lot of human time. Because it's not that it has the potential to reshape us. It will, unless we make more time. And she's very explicit in the article about this. What we have to invest in is literally more time. Give educators more time with students. Smaller class sizes. More connection. More opportunities for relationality. More opportunity with less to do. Less directives, less paperwork, more hanging out, talking, figuring things out together.

Isabelle, we need you on the podcast. You're amazing. Love your work. Love what you're doing.

01:46:24

I want to leave us on a high note, because we've been pretty dark.

What gives me hope is honestly being an educator. Working with students. Because they are the future, and no matter what, you cannot force an entire generation to go against what it wants. And everything I see from this generation is relational, and is moving in the direction of a simpler time, not in the way of shunning technology, not Luddites in any way. They use it to create beautifully. They're not recreating Thoreau. But they really need to think about how they're using their time. They really need to make time for each other. Vinyl parties. Nightclubs with phone lockers. Just be here, right now, be here. I think we'll be okay.

The world is still good and still beautiful.

01:48:19

I love NotebookLM. Me too.

Do you know I have a document coming? I might have gone down an ADHD waggle hole on this one. Jake, what are you doing right now?

I am editing a document I'm going to use as a lead magnet for when people subscribe to my email. It's about 20 pages right now of different ideas of how to use NotebookLM by role in education, different age groups, administrative assistants, management, operational IT, all of it. It has the ability to be truly incredible.

Want to do that on the next one?

No. The next one is Legoland. That's right. Jake is headed to Legoland Florida Resort tomorrow. We're having an EduProtocols Academy spread throughout the whole park, with 20-minute passing periods between sessions so you can go on a ride. Full day pass inside the park. The first day ends at noon so people have the rest of the day to play, and the second day ends around three.

Talk about relationality.

I know. I can't wait. There will be pictures and an episode.

I know. I wish you were coming.

Get a plane ticket.

I can't, I'm sorry.

Whatever. All right. Scene.

What Teachers Have to Say is hosted by Jake Carr and Nathan Collins.