⚖️ Teacher's Workbook: ADHD + AI — The Honest Case

Library class 3 of 3 · 90 minutes · 21 slides · pairs with the projector deck

This is the debate class. Classes 1 and 2 teach people to do something; this one teaches them to think clearly about the whole thing — four real pros, five real cons, and the fence that sits between an ADHD brain and a machine built out of everything it craves. Your job all night is to keep the seesaw honest: never let a pro land without its edge, never let a con land without its counter-move. Each segment below matches one slide and gives you the goal, an expanded version of what to say, what to do, and how the moment usually goes sideways. Skim it the night before, print the worksheet, and teach off the deck.

Open the projector deck → ← All classes

THE SHAPE OF ITWhat this class does

People walk in either hyped about AI or braced against it. They walk out with something better than either: a working scale. Four pros they can name, five cons they can name, one rule to weigh any AI moment against, and a ten-minute exercise already run on their own phone. By the end, every person in the room should be able to:

What this class deliberately does not do: pick a winner. AI does not cure, diagnose, or treat ADHD, and no study shows any AI tool improves ADHD outcomes — you'll say that plainly, twice. But the cons don't win either: the diary-study voice, the coaches' small daily uses, and the room's own emergency-prompt results are real. The verdict slide says it best: power tool, not a miracle. Hold both, all night.

BEFORE YOU TEACHMaterials & setup

Bring:

Room: tables beat rows — people will be on phones and writing in the worksheet table. Wifi password written large and visible before you start. Recruit one helper for the ten-minute exercise; twenty phones hit twenty different snags and you can't reach them all alone. This class is the most talk-heavy of the three — one hands-on block instead of two sprints — so your energy management matters more here. The pros/cons alternation is your friend: the structure itself keeps attention resetting every five minutes.

Know your controls: arrow keys or clicking advances. Bullet slides reveal one line per press — you control the pace. The corner badge shows the run-sheet time so you always know if you're ahead or behind. The exercise slide has a 10-minute timer with a 3-2-1 count-in, and it keeps counting even if you bump a key. If people arrive who took Class 1, great — they'll recognize the four words at 1:15 and you can let them say it first.

AT A GLANCEThe run sheet

ClockSegmentWhat's happening
0:00WelcomeFrame the debate: pros, cons, and the fence between
0:04The one ruleThe sentence the whole class hangs on
0:09Pro 1: thought-speed captureThe GTD capture rule — talking is instant
0:14Pro 2: zero judgment, any hourThe diary-study voice — a listener that never sighs
0:19Pro 3: made of what ADHD cravesThe 2018 novelty finding — and its shadow
0:24Pro 4: borrowed structureBody doubling — sworn by, never proven
0:29Con 1: the rabbit holeThe same circuit, pointed the wrong way
0:34Con 2: walls of textThe essay problem — SMALLER and SHORTER
0:39Con 3: nothing is provenZero studies — said plainly, mid-class
0:43Con 4: confidently wrongGoblin Tools' disclaimer as the honest standard
0:47Con 5: a stranger with a great memoryPostcard rule installed right before phones come out
0:52Exercise (10 min)Everyone runs the emergency prompt live
1:04Share-outWhat came back? Sideways beats smooth
1:08Pop quizThe honest state of 2026, retrieved not recited
1:12The verdictPower tool. Not a miracle.
1:15The method betweenYAP · MAP · TAP · SNAP as the fence
1:17Climbing out of the rabbit holePark it · SNAP first then dive · the 20-minute question
1:18What real coaches doFive coaches, five small boring uses
1:23Notice what that list is notNobody's life is run by AI
1:26The all-in entrepreneurPeter Shankman — the other end of the dial
1:29Try it tonightOne 60-second yap, AI just listens

SLIDE BY SLIDEThe segment guide

0:00 ADHD + AI: The Honest Case

Goal: frame the night as a fair trial, not a pitch — and earn both the skeptics and the enthusiasts in the first two minutes.

Say

"Welcome to the argument class. Some of you walked in thinking AI is going to change your life, and some of you walked in thinking it's a scam with good marketing — and here's my promise: both of you are going to hear your side taken seriously tonight. Four real pros. Five real cons. Yes, the cons outnumber the pros — count them yourself as we go, and notice I'm not hiding that. I've spent twenty-nine years behind a bar listening to people, and the one thing I know is that nobody trusts a person who's only ever excited. So tonight nothing gets sold. Every claim I make stays small enough to check. And if you leave remembering exactly one sentence from the whole ninety minutes, it's the next slide — so don't get comfortable yet."

Do

Ask for a quick show of hands — "who's excited about AI? who's suspicious?" — and genuinely thank both groups. You've just told the room the debate is real, and you've located your skeptics for later. Note where they sit.

Watch for

People who took Class 1 or 2 expecting another hands-on night. Set the shape early: "Tonight is one exercise and a lot of honesty — it's the thinking class of the three."

0:04 The one rule

Goal: install the measuring stick before any evidence arrives — every pro and con tonight gets weighed against this sentence.

Say

"Here it is, the sentence: if AI doesn't make your next step smaller — it's not helping. Say it with me once, because you'll need it in an hour when I can't be there: if AI doesn't make your next step smaller, it's not helping. You've been sold miracle planners before. Every one of you owns a system that was going to fix everything — the binder, the app, the color-coded calendar. This class is not that, and this rule is why. It doesn't ask whether the AI is impressive, or smart, or fun to talk to — it can be all three and still be eating your evening. It asks one thing: is my next step smaller than it was ten minutes ago? Every pro tonight passes that test sometimes. Every con is what happens when it fails. Keep score."

Do

The rule reveals in two lines — press for the first half, pause, press for the punch. Have the room say it back once. Point at the worksheet: it's printed at the top, and there's a pros/cons table they'll fill in as the night goes.

Watch for

Someone who wants to debate the rule itself ("what about brainstorming? what about company?"). Good instinct — defer it: "Hold that. The coaches at the end of tonight are going to answer you better than I can."

0:09 Pro: it catches thoughts at thought speed

Goal: land the first pro on solid, pre-AI ground — the capture insight is decades old; AI just made it instant. Then attach its edge before moving on.

Say

"Pro number one, and it starts before AI existed. The Getting Things Done people figured out a rule a long time ago: if capture isn't instant, it won't get done. The thought that takes four steps to write down is a thought that's already gone — and for an ADHD brain, 'gone' happens in seconds, not minutes. So what's the most instant capture there is? Talking. You can talk faster than you can type, faster than you can find the notes app, faster than you can find a pen. That's the entire reason this whole method starts with yapping. Now, the edge on this pro, because every pro tonight gets one: catching a thought is not the same as doing anything about it. A perfectly captured thought that never gets smaller is just a very well-organized pile. Capture is the doorway — the rule still decides what's helping."

Do

Have them write pro 1 into the worksheet table in their own words — a few seconds of silence while pens move. You'll repeat this beat after every pro and con; it's how the table becomes theirs instead of yours.

Watch for

The enthusiasts nodding too fast. They already believe the pros; your edges are aimed at them. Make eye contact with that side of the room when you deliver the "well-organized pile" line.

0:14 Pro: zero judgment, any hour

Goal: let a real ADHD adult's own words carry this pro — it's the most emotionally credible evidence of the night. Then name the shadow inside it.

Say

"Pro two, and this one I'll let someone else say. In a peer-reviewed diary study, an ADHD adult described brain-dumping to AI like this: in two seconds I can scratch that itch, get it out, and go back to what I need to do. Sit with that. Two seconds. No 'have you tried a planner.' No sigh. No face — you all know the face — from the person you've asked to listen one too many times. A listener that never sighs, at 2am, on the worst day, for free. If you've spent decades editing yourself before you speak, you know exactly why this matters. And here's the edge, and I want you to hold onto it for fifteen more minutes: a listener that never sighs is also a listener that never says 'okay, that's enough now, go do the thing.' The patience is the gift. The patience is also the trap. Con number one is coming for exactly this."

Do

Deliver the quote slowly and let it breathe — this is the slide where the skeptics' arms uncross. Then pens to the table again: pro 2, their own words.

Watch for

Someone visibly moved — "a listener that never sighs" lands hard on people who've worn out their human listeners. Don't spotlight; slow your pace ten seconds and keep the warmth on.

0:19 Pro: it's made of what ADHD brains crave

Goal: the science-flavored pro — deliver it with the exact hedged language the research deserves, because this same finding becomes con 1 in ten minutes.

Say

"Pro three, and watch my wording, because I'm going to be careful on purpose. A 2018 brain-imaging study suggests — suggests, not proves — that ADHD brains weight novelty about twice as heavily as other brains. Twice. Now think about every productivity tool you've ever been handed. The planner, the timer, the filing system — every single one asked that brain to suppress the craving. Sit still, ignore the shiny thing, do the boring thing. And every single one lost, because you don't out-discipline your own wiring. AI is the first tool built out of the craving instead of against it: it's novel every time you open it, it answers instantly, it never judges. That's why it feels different from every app that came before. And the edge — say it before I do, some of you see it coming: a tool made of exactly what your brain craves is also the perfect bait. Hold that thought. Ten minutes."

Do

Three lines, one per press — the finding, the old tools, the new tool. Lean on the word "suggests" both times; you're modeling how to hold research loosely. Table: pro 3.

Watch for

Someone hearing "2x novelty" as proof AI is good for ADHD. It isn't — it explains the appeal, not the outcome. If it comes up, say exactly that: "It tells us why it feels good. Whether it helps is a different question, and slide nine is honest about it."

0:24 Pro: borrowed structure on demand

Goal: the last pro, wearing its honest label in the same breath — this slide is where the room learns you'll caveat your own side.

Say

"Last pro. People with ADHD have sworn by body doubling for years — working alongside another person makes starting easier, somehow. And AI can simulate a version of that presence: someone to check in with, restart cards, one question at a time instead of a blank page, at 2am when no human is available and no human should be. Borrowed structure, on demand. Now here's the label, and notice I'm putting it on my own side of the argument, not waiting for you to catch me: loved by thousands, not clinically proven. Thousands of people swear by body doubling. The formal research barely exists. Nobody — including me, especially me — gets to stand up here and call it proven. What you get to do is run the experiment on yourself, which you'll do in half an hour. So that's the case for: instant capture, a listener that never sighs, a tool made of the craving, and borrowed structure. Four pros, all real, all edged. Now the other side of the scale."

Do

Table: pro 4. Then pause and mark the turn physically — step to the other side of the stage, take a breath, change your posture. The room should feel the seesaw tip before you say a word of the cons.

Watch for

The skeptics warming up as the cons approach — they've been waiting. Good. Your job on the next five slides is the mirror of the last four: every con gets its counter-move, and you'll aim those at the skeptics the way you aimed the edges at the enthusiasts.

0:29 Con: the ADHD rabbit hole

Goal: the biggest con of the night, delivered with full weight — and the promise of a counter-move, so it warns without paralyzing.

Say

"Con number one, and it's the one I warned you about twice. The exact novelty circuit that makes AI feel wonderful — the one from the 2018 study — is the one researchers tie to chatbot rabbit holes. Same wiring. Same craving. One tab becomes forty minutes. You opened it to plan your Tuesday and now you're deep in a conversation about something fascinating that is not your Tuesday. And here's the cruelest part: endless prompting feels productive. You're typing, it's answering, ideas are happening — and the task didn't get one minute smaller. Three hours of talking about the thing is three hours the thing didn't get. Run the rule: did my next step get smaller? No? Then that wasn't help, that was a very sophisticated slot machine. The counter-move is not willpower — you'd have used willpower by now if it worked. It's a fence between the novelty and your next action, and I'll hand you the whole fence at the end of tonight. For now just write the con down honestly."

Do

Table: con 1. Deliver this one straight at the enthusiast side of the room — it's their con. Don't rush past "feels productive and isn't"; repeat it if the room is nodding.

Watch for

Skeptics feeling vindicated and starting to check out — "knew it, case closed." Pull them back by name-checking the balance: "Notice this con doesn't erase the diary-study lady. Both are true at once. That's the whole class."

0:34 Con: walls of text re-overwhelm

Goal: the most fixable con — teach the two-word counter-move in the same breath, because they'll need it during the exercise in twenty minutes.

Say

"Con two. Ask an AI anything and its default answer is an essay. A nine-step plan with sub-bullets. And here's the problem nobody at the AI companies apparently thought about: an essay is a new pile of crumpled papers for the exact brain it was supposed to help. You came in overwhelmed, you asked for help, and it handed you more to read. Chest tightens, tab closes, and now you feel bad about the AI too. Real con — write it down. But this is also the most fixable con on the list, and the fix is two words the machine has to obey. Plan too big? Say SMALLER — it has to shrink the step, and you can say it five times in a row, no shame. Wall of text? Say SHORTER — it has to cut the words. You're not asking a favor; you're giving an instruction, and instructions are the one thing these things always take. You will personally use SHORTER in about twenty minutes. I'd bet on it."

Do

Table: con 2 — and have them write SMALLER and SHORTER next to it. These two words are the ammunition for the exercise; make sure they're on paper before phones come out.

Watch for

"Does that work on any AI?" — yes: "Any chat AI. It's not a feature, it's an instruction." Class 1 alumni will know this beat; let one of them answer.

0:39 Con: nothing is proven yet

Goal: the credibility spine of the whole class. Delivered flat, mid-class, unprompted — this is the slide that separates you from every AI pitch they've heard.

Say

"Con three, and I need you to hear this one in my plainest voice. Zero studies show any AI tool improves ADHD outcomes. Zero. Not this method, not any app, not any of it. If anyone — on a stage, in an ad, in your feed — promises that AI will fix your ADHD, they are selling something, and now you know it for a fact. So why am I standing here? Because of what we do get to say, and notice how small it is: AI can support. It can reduce load. It can help you start. Those are the biggest claims I'll make tonight, and every one is checkable on your own phone in your own week. Here's the counter-move on this con: unproven doesn't mean useless — it means the burden of proof is on your own experiment, not on anyone's promises. Track your week. Count what got smaller. Your data beats my claims, and it definitely beats an ad's."

Do

Slow down and go flat — contrast is what makes it credible. Table: con 3, and suggest they write the word "zero" in it. This exact fact is the pop-quiz answer at 1:08; you're planting it now.

Watch for

Relief on the skeptic side and confusion on the enthusiast side — "then why are we here?" is a fair face to see. The "can support, can reduce load, can help you start" triplet is your answer; say it twice if needed.

0:43 Con: it's confidently wrong sometimes

Goal: make "AI can be wrong" concrete using the one company honest enough to admit it — and give the room a scaling rule for how much to care.

Say

"Con four. AI is sometimes wrong — and it's wrong the worst possible way: confidently, politely, in perfect grammar. It doesn't stammer when it's unsure. It sounds exactly as certain about nonsense as about facts. Now, most companies bury that in fine print. One tiny ADHD tool, Goblin Tools, puts it right up front — the most honest disclaimer in the industry: nothing returned should be taken as truth, only guesswork. I love that sentence. I want you to treat every AI — every single one — the way that one little tool asks to be treated. Guesswork until checked. Helper, not doctor. The counter-move is a scaling rule: the more a wrong answer would cost you, the less you let AI answer it alone. A clumsy suggestion for a tiny task? Costs you nothing; you're the editor. Anything medical, legal, or money? That goes to a professional, full stop. Tonight's exercise is deliberately at the cheap end of that scale."

Do

Table: con 4. The Goblin Tools line is quotable — give the room a second to write it down; some will want it verbatim.

Watch for

"How would I even know it's wrong?" — honest answer: for tonight's low-stakes uses, being wrong barely costs anything, and that's by design. The rule scales with stakes. Don't pretend there's a detection trick; the protection is choosing what you ask.

0:47 Con: it's a stranger with a great memory

Goal: the last con doubles as the safety briefing — the postcard rule must be installed NOW, because phones come out on the very next slide.

Say

"Last con, and it's placed here on purpose, because in about ninety seconds every phone in this room comes out. The AI chat window is not your diary, and it's not your doctor's office. It's a stranger with a great memory. So the privacy rule, always, everywhere, tonight and forever: nothing goes into an AI that you wouldn't write on the back of a postcard. Fake names only — Person A, Company B, Project X. No medical details, no legal situation, no account numbers, no money specifics. ADHD coach Paula Engebretson — a professional who uses these tools every day — puts her own line at: nothing I wouldn't be comfortable having anyone else see. That's the standard from someone who actually uses it, not someone scared of it. Follow the postcard rule and the scary privacy question mostly answers itself, because there's nothing sensitive in there to leak. Say the rule's name back to me — what rule? Postcard rule. Good. Phones out."

Do

Table: con 5 — then get the call-and-response on "postcard rule" before advancing. Point at the postcard-rule box on the worksheet. Do not skip or rush this slide to save time; it is the consent form for the exercise.

Watch for

Anyone already typing before you finish. Kindly stop the room: "Thirty more seconds and then everyone at once — I need the rule to land first." The eager typer at 0:47 is the oversharer at 0:55.

0:52 Try it: the emergency prompt ⏱ 10:00

Goal: the centerpiece. Every person runs one real prompt on their own phone, gets an answer, and — critically — many of them get to use SHORTER on a too-long reply. The class's whole argument becomes personal experience here.

Say

"Phones out, ten minutes, and this is the whole hands-on portion of tonight, so let's make it count. Open any free AI app — whichever one you have. You're going to type it exactly this: I'm overwhelmed right now. Don't ask me questions. Tell me one tiny thing to do in the next 5 minutes. Two sentences max. It's printed on your worksheet word for word — copy it. This is the prompt for your worst day: the day you can't think, can't sort, can't even yap. Notice what it does: it forbids questions, it forces tiny, it caps the length. And here's my prediction — some of your AIs are going to answer with an essay anyway, because that's the wall-of-text con happening live. If yours does, do not apologize to it, do not read the essay. Type one word: SHORTER. Then watch it obey. That moment — the machine folding to a one-word instruction — is the fence, working. Postcard rule applies as always. Go."

Do

Start the slide timer — it has a 3-2-1 count-in and runs the full ten minutes even if you bump a key. Then WALK, you and your helper covering opposite sides, the entire time. First three minutes: getting everyone in an app at all — pair anyone with a dead or stubborn phone with a neighbor; shared phones work fine. Middle minutes: read screens over shoulders. The three fixes you'll give most, in order: (1) the answer came back long — stand next to them while they type SHORTER, and celebrate quietly when it obeys, then tell them to say SHORTER again if it's still long; (2) they've softened the prompt into a polite question — have them retype it verbatim from the worksheet, the bluntness is load-bearing; (3) they got a good answer and immediately started chatting with the AI about it — point at the one rule on the worksheet and grin: "That's the rabbit hole, live. You have your tiny thing. You're done." Last two minutes: tell finishers to write the answer they got into the worksheet and note whether SHORTER was needed. Run your own emergency prompt on your phone too — visibly. If venue wifi dies, hotspot from your phone and run yours projected while pairs share working phones.

Watch for

The essay-anyway case is the jackpot, not the failure — flag it for the share-out: "Don't fix it quietly, tell us about it in five minutes." Someone typing something clearly sensitive into the prompt (real names, medical situation): quiet, kind, private — "postcard version of that one" — never call it out to the room. The person whose AI asked a question anyway despite the prompt: "It disobeyed — good catch. That's con four, confidently doing its own thing. Say it again, firmer." The person frozen at the blank box because even the pre-written prompt feels like a decision: sit down next to them and read it out while they type. And the fast finisher going down a rabbit hole with their new AI friend: the exercise is the lesson; land them.

1:04 What came back?

Goal: harvest the exercise publicly — especially the failures, because the room learns the cons best from a neighbor's screen, not your slides.

Say

"A few volunteers — did your AI behave? What did it tell you to do? And most importantly: whose AI ignored the rules and sent an essay anyway? Hands up — look around, you were not alone. Did SHORTER work when you said it? That's the single most useful thing tonight can prove: the machine's defaults don't fit your brain, and one blunt word bends the machine to fit anyway. And who got told to do something genuinely tiny and thought — huh, I could actually do that? Notice you didn't need a nine-step plan. You needed one small true thing and permission to ignore everything else. The room learns more from what went sideways than what went smoothly, so if yours was weird, wrong, or stubborn — that's not an embarrassing result, that's tonight's syllabus happening on your phone."

Do

Count to five in your head after asking — someone always breaks the silence. Take the flagged essay-anyway person first if they're willing. Thank every share by name. If nobody volunteers, show YOUR result projected and ask "who got something like this?" — hands count as sharing.

Watch for

An overshare — the emergency prompt was written from real overwhelm, and share-outs can slide into the actual crisis behind it. Catch it kindly: "That's real, and it's exactly what the postcard rule protects — thank you for trusting us." Then move on; this is not group therapy. Also watch the person whose AI gave a genuinely bad suggestion: "You're the editor. It's guesswork until you approve it — remember the little goblin."

1:08 Pop quiz

Goal: retrieval practice on the single most important fact of the night — the room commits before the reveal.

Say

"Pop quiz — which of these is actually true in 2026? Option one: AI is clinically proven to treat ADHD. Option two: no study shows any AI tool improves ADHD outcomes. Option three: body doubling is clinically proven. Shout it, point at it, commit — no fair waiting for me." [Let the room answer before revealing.] "Two. The honest one. Option one is what the ads want you to believe — nobody gets to claim it. Option three is sneaky, because body doubling is real in people's lives and thousands swear by it — we said so ourselves an hour ago — but sworn-by and proven are different words, and tonight you learned to hear the difference. That's the honest state of 2026: the tools can still help you start, and nobody gets to promise outcomes yet. If you can hold both of those at once, you're now better calibrated than most of the internet."

Do

The quiz slide takes the room's commitment before the reveal — use it, and let wrong guesses be gold: option three catches Class 1 alumni who remember the body-doubling slide fondly. "Why is three wrong?" is the best discussion of the night if someone bites.

Watch for

Everyone nailing it instantly — celebrate and bank the minutes; the closing run from 1:12 always wants more room than it gets.

1:12 The verdict

Goal: collapse ninety minutes of seesaw into four words the room can carry out the door — without letting either camp claim victory.

Say

"So — verdict time. You've heard four pros. They're real: I watched some of you get a genuinely useful answer twenty minutes ago. You've heard five cons. They're real too: some of you got the essay, some of you felt the rabbit-hole tug in a ten-minute exercise. So which side wins? Neither. Here's the verdict: power tool. Not a miracle. A power saw is a genuinely great tool — and nobody hands you one without a guard, and nobody calls it a carpentry cure. The pros are real. So are the cons. And the difference between someone the pros help and someone the cons eat is not intelligence, not discipline, not tech skills — it's whether a method sits between you and the machine. A fence. Which raises the obvious question: what fence? Some of you already know the answer, and it's four words long."

Do

The verdict lands in two reveals — "Power tool." then "Not a miracle." Pause between them. This is the sentence for the worksheet's verdict line; give pens a moment.

Watch for

Either camp declaring victory — the enthusiast hears "power tool!" and the skeptic hears "not a miracle!" Both are half-listening. The load-bearing sentence is the third one, about the method between; make sure it lands on both sides.

1:15 The method that sits between

Goal: name the fence — for alumni it's a reunion, for newcomers it's a pointer to Class 1. Keep it brisk; this is a signpost, not a lesson.

Say

"Four words. Some of you know them — say them with me. YAP: dump the backpack. Talk messy, out loud, no organizing while you dump. MAP: the AI sorts your pile into piles you can look at. TAP: you pick up one paper — one action, small enough to actually do. SNAP: you save your game before you stop, so tomorrow starts warm. Now look at what those four words do to tonight's argument. Every pro gets used: the capture, the judgment-free listener, the novelty, the borrowed structure — all four are in there. And every con gets fenced: the yap has an end, the map forces SMALLER, the tap points at real life instead of another prompt, and the snap gets you out clean. That's what 'a method between you and the machine' means. It's a whole class of its own — class one of this series, and it's free, and it'll still be free next month. Tonight you just needed to know the fence exists."

Do

Four words, one per press, room echoing — alumni will carry the chant and newcomers borrow their confidence. Then move; resist the urge to teach Class 1 in three minutes. Point to lessons.html on the site for the full class.

Watch for

Newcomers looking lost at the sudden vocabulary. One reassurance line: "If those words are new, that's the other free class — tonight's job was just the verdict."

1:17 Climbing out of the rabbit hole, thread in hand

Goal: pay off the debt from con 1 — three concrete brake-style moves for the night the rabbit hole wins. This is the most practically valuable slide of the closing run.

Say

"At 0:29 I promised you a fence for the rabbit hole. Here it is — three moves, because willpower isn't one. First, understand the grip: your brain clamps onto the shiny thought because it's afraid of losing it. That's not weakness, that's your capture system doing its job too hard. So — move one, park it: one line in an Ideas pile, and the grip releases on its own. You don't fight the thought; you promise it a parking spot. Move two, for when you genuinely want the dive: SNAP first, then dive. Save your place — four lines — take the ten minutes guilt-free, and come back for free, because your spot is held. Move three, the 20-minute question, for when you're already down there: is this making my next step smaller? Ask it honestly. Flow passes that test. The spiral doesn't. Three brakes: park it, save-then-dive, and the question. Write all three down — they're on the worksheet, and you'll need them by Thursday."

Do

Four lines reveal one per press — grip, park, snap-then-dive, question. Point at the worksheet's rabbit-hole box. This slide and the next share a tight two-minute window on the clock, so keep the delivery crisp; the depth is on the worksheet.

Watch for

"But sometimes the dive IS the good part" — agree! That's exactly what move two is for: the dive taken on purpose, with your place saved, is flow. The dive that takes you is the spiral. The question at minute 20 is how you tell them apart from the inside.

1:18 What real ADHD coaches actually do with AI

Goal: ground the whole debate in five real professionals' actual daily habits — evidence of use, deliberately not evidence of outcomes.

Say

"Let's end with what the people who do this for a living actually do — real ADHD coaches, real habits, and notice how unglamorous every single one is. Paula Engebretson — the postcard-rule coach from earlier — uses AI for analogies to learn faster, and to check the tone of her emails. William Curb plans his day with the ChatGPT app at breakfast. Marla Cummins writes the blunt draft herself — her words, her thinking — and lets AI soften it. Tara McGillicuddy drafts replies to charged emails, not to send them, but to create a pause between the feeling and the reply. And Lynne Edris uses it to beat blank-page syndrome — it gets her moving. Five professionals. Five uses you could copy tomorrow morning with zero technical skill. Look at your worksheet as I read them — there's a line asking which one you'd steal. Pick one now, while they're fresh."

Do

Five lines, one coach per press — brisk, warm, no dwelling. The "which would you steal?" worksheet line does the real work; give ten seconds of pen time at the end.

Watch for

The clock — 1:18 to 1:23 must fit this slide and the pen moment. If you're behind, trim your commentary, never the five names; the list is the payload.

1:23 Notice what that list is not

Goal: the anti-hype punchline — the room realizes the experts' secret is smallness, which quietly validates the one rule.

Say

"Now notice what that list is not. Not one coach said AI runs their life. Nobody said it planned their year, managed their business, or fixed their ADHD. Analogies. Breakfast planning. Softening a draft. Pausing before a spicy email. Getting off the blank page. Small, boring, daily uses — that's what working actually looks like. If you came tonight expecting the experts to have some advanced setup you couldn't handle, here's your relief: they don't. The ceiling of expert use is within reach of everyone in this room, tonight. And one more piece of honesty from inside the industry: William Curb — a man who reviews ADHD tools for a living — admits he skips most of the tools he reviews. The man drowning in options uses almost none of them. The tool matters that little. The habit of keeping it small matters that much. Which is the one rule again, wearing five coaches as evidence."

Do

Slow, warm delivery — this is the reassurance slide for the 50+ room that suspects everyone else is more advanced. Land the Curb admission with a smile; it's the funniest true thing in the deck.

Watch for

The relief is the tell — shoulders dropping when they hear "small, boring, daily." If the room is with you, add nothing; over-explaining flattens this one.

1:26 One more: the entrepreneur who went all-in

Goal: show the far end of the dial honestly — heavy use exists and works for one specific, named person — without letting it become the room's homework.

Say

"Fairness demands one more story, because the coaches were all cautious and the dial does go further. Peter Shankman — entrepreneur, host of the Faster Than Normal podcast — went all-in. His words: I have massive ADHD... every minute I waste on a decision I could have automated is a minute stolen from something that matters. So ChatGPT drafts the first version of anything he's dreading — all I have to do is edit that, he says. Claude pressure-tests his ideas before clients ever see them. Beautiful.ai builds his slides. Two companies, a podcast, keynote speaking, a kid, a dog — and his summary is: AI is how I do it. That's real, and he's thriving. Now the edge, same as every pro tonight got: that's one person's dial setting, not a prescription. Notice even his all-in version obeys the rule — draft the dreaded thing, edit instead of start. Everything he automates makes his next step smaller. Start where the coaches are. The dial will still be there."

Do

Lines reveal one per press. Let Shankman's energy contrast with the coaches' calm — the room should feel the dial turn. Then pull it back to the rule in your own steady voice.

Watch for

The enthusiast hearing this as the real recommendation and the coach slides as the warm-up. Say it plainly if you see it: "He built to that over years, on top of knowing exactly what he's dreading. Day one is a sixty-second yap — which is conveniently the next slide."

1:29 Try it tonight

Goal: shrink the entire debate to one sixty-second assignment — small enough that skipping it is harder than doing it — and close warm.

Say

"Ninety minutes of pros and cons, and here's the entire homework: tonight, one sixty-second yap into any free AI app. Tell it: just listen, don't solve, until I say DONE. Then talk for one minute about your day — messy, unorganized, postcard rule as always — and say DONE. That's it. That's the whole assignment. Notice what it is: it's the smallest possible version of the biggest pro — the listener that never sighs — with a fence already built in, because DONE is an ending and endings are what rabbit holes don't have. You're not adopting AI tonight. You're not buying anything, and you're definitely not fixing your ADHD, because nothing does that yet and now you know it. You're running one sixty-second experiment, and your own Tuesday is the data. That's the honest case: power tool, not a miracle, and the fence is four words long. Thank you — everything from tonight lives free at adhdyap.com."

Do

The slide celebrates as it lands — let the confetti fire, lead the applause, then STOP TALKING. The try-it-tonight line is on the worksheet; tap it as your last gesture. Stay 15 minutes — this class's one-on-one questions skew personal and are the realest of the series.

Watch for

The librarian — thank them by name in front of the room; that's how the next booking happens. And catch your skeptics on the way out: "So — where'd you land?" Their answer tells you whether the seesaw stayed honest.

BE READYTough questions from the room

"If zero studies show it helps, why are you teaching a class on it?"

"Because 'unproven' and 'useless' are different words, and the space between them is exactly where honesty matters most. Millions of people with ADHD are already using these tools — the only question is whether they use them with the cons named and a fence in place, or off a hype video. What I get to claim is small: can support, can reduce load, can help you start — and you ran that experiment yourself tonight in ten minutes. What nobody gets to claim is outcomes. A class that teaches both halves is, as far as I can tell, the rarest thing in AI right now."

"Doesn't that 2x novelty study prove AI is good for ADHD brains?"

"No — and being careful here is the whole class in miniature. The 2018 imaging work suggests ADHD brains weight novelty about twice as heavily. That explains why AI feels so good to a brain like ours — and the very same circuit is the one tied to chatbot rabbit holes. One finding, both edges. It tells us about the pull, not the payoff. Whether the pull helps you depends entirely on whether a method sits between the novelty and your next action."

"Those five coaches use it every day — isn't that proof it works?"

"It's proof of use, not proof of outcomes — and the coaches themselves would be the first to say so. Five professionals finding small daily value is genuinely meaningful evidence about where the value lives: starting, drafting, pausing, softening. But sworn-by and proven are different words — same as body doubling, which thousands swear by while the formal research barely exists. Copy their habits cheaply, track your own week, and let your data make the claim."

"My emergency prompt came back with an essay anyway. Is my AI broken?"

"Not broken — default. An essay is what these things do when nobody stops them, which is exactly con two happening on your phone. The fix is one word: SHORTER. Say it, watch it obey, and say it again if the answer is still long — there's no limit and no politeness required. If it asked you a question despite being told not to, same story: repeat the instruction, firmer. You're not asking favors; you're giving instructions, and instructions are the one thing they always take."

"Which AI app should I use?"

"Any free chat AI does everything we did tonight — the emergency prompt, SHORTER, the sixty-second yap. Use whichever is already on your phone, stay on the free tier, and don't pay anyone a dollar until small daily use is a habit. Remember William Curb: the man reviews ADHD tools for a living and skips most of them. The tool matters that little. The Tools page on the site keeps an honestly-labeled list with prices if you want to compare later."

"Is my information safe with these AI companies?"

"Treat every AI like what it is — a stranger with a great memory — and let the postcard rule do the protecting: fake names, nothing you wouldn't write on a postcard, no medical, legal, or money details. Paula Engebretson's version of the line is worth stealing: nothing she wouldn't be comfortable having anyone else see. Follow that and the privacy question mostly answers itself, because there's nothing sensitive in there to worry about. I can't vouch for any company's policy, and neither can anyone else giving a talk."

"Can AI diagnose me? I've always suspected I have ADHD."

"No — not with your whole life story pasted in, and don't let it try. It doesn't diagnose, treat, or cure ADHD; it's guesswork until checked, remember, and a diagnosis is the last place you want guesswork. What it CAN do is help you prepare for the real appointment: yap about how things have actually been going, have it sort your mess into a one-page list of things to mention. That list goes to a clinician. The answer comes from a clinician. Never from a chatbot."

"I already lose hours to my phone. Won't adding AI make it worse?"

"You might be the person this class was most honestly built for — that instinct is con one, and it's real: same novelty circuit, and endless prompting feels productive while your task gets nothing. So don't add AI bare; add it fenced. Everything we gave you tonight has an ending built in: the emergency prompt caps itself at two sentences, the sixty-second yap ends at DONE, and the three rabbit-hole brakes — park it, SNAP first then dive, the 20-minute question — are on your worksheet. If after a week the fence isn't holding, believe your own data and step back. That's not failing the class; that IS the class."

PRINT MEStudent worksheet — ADHD + AI: The Honest Case

Print one per seat (this section prints clean on its own — use the print button up top). It's the scale, the fence, and the homework in one page.

⚖️ ADHD + AI: The Honest Case — my worksheet

Name: ____________________ · Date: ____________

📏 The one rule (memorize this one)

If AI doesn't make your next step smaller, it's not helping.

The scale — fill each in, in your own words, as we go

The claimIn my own words
Pro 1Catches thoughts at thought speed (instant capture)______________________________
Pro 2Zero judgment, any hour (the listener that never sighs)______________________________
Pro 3Made of what ADHD brains crave (novelty)______________________________
Pro 4Borrowed structure on demand (sworn by, not proven)______________________________
Con 1The ADHD rabbit hole (feels productive, isn't)______________________________
Con 2Walls of text re-overwhelm (fix: SMALLER · SHORTER)______________________________
Con 3Nothing is proven yet (zero studies on outcomes)______________________________
Con 4Confidently wrong sometimes (guesswork until checked)______________________________
Con 5A stranger with a great memory (postcard rule)______________________________

🧪 The emergency prompt (copy it EXACTLY)

"I'm overwhelmed right now. Don't ask me questions. Tell me one tiny thing to do in the next 5 minutes. Two sentences max."

If the answer comes back long, say SHORTER — and again if it's still long.

What my AI told me to do:

Did I need SHORTER?   ☐ yes   ☐ no   ☐ twice

🐇 Climbing out of the rabbit hole (three brakes)

My brain grips the shiny thought because it's afraid of losing it. So:

1. Park it — one line in an Ideas pile, and the grip releases on its own.

2. SNAP first, then dive — save my place (four lines), take the 10 minutes guilt-free, return for free.

3. The 20-minute question — is this making my next step smaller? Flow passes. The spiral doesn't.

📮 The postcard rule (the one safety rule)

Nothing goes into an AI that I wouldn't write on a postcard. Fake names only — Person A, Company B, Project X. No medical, legal, or money details. And: AI can be wrong. Guesswork until checked. Helper, not doctor.

🧑‍🏫 The coach habit I'd steal (circle or write one)

Analogies to learn faster · plan the day at breakfast · write blunt then let AI soften · draft the charged reply to create a pause · beat the blank page

🧰 The verdict, in my words

Power tool. Not a miracle. The difference is a method between me and the machine: YAP · MAP · TAP · SNAP

🌱 Try it tonight

One 60-second yap into any free AI app: "Just listen, don't solve, until I say DONE." Talk for a minute. Say DONE. That's the whole assignment.

Everything from tonight, free: adhdyap.com · AI supports an ADHD brain. It does not cure, diagnose, or treat ADHD — and it can be wrong.

AFTERWARDAfter class