🔬 Teacher's Workbook: The Evidence Files

0:00 The Evidence Files

Goal: promise receipts, and promise the skill transfers beyond tonight.

"Quick poll, hands up: who here has bought a planner, an app, a supplement, or a course that promised to fix your focus? Keep it up if it's still fixing your focus. Yeah. That's this class. For the next ninety minutes we're going to do something no ad will ever do for you — we're going to sort everything you've heard about ADHD and AI into three piles: what's actually proven, what's promising but thin, and what's just for sale. And you're not going to have to take my word for anything, because the real thing I'm teaching tonight is the sorting itself. You'll run it on a live claim with your own phone before you leave."

The hands that stay up on the second question, ironically — those people are your best volunteers later. They've paid the tuition already.

0:04 Why a whole class about studies

Goal: name the predator without shaming the prey.

"Here's the uncomfortable economics: ADHD brains are a target market. We're hopeful, we're impulsive on a bad day, and we are actively looking for the thing that finally works. The ads know exactly where it hurts. So this is not a class about being smarter than the people who bought something — I've bought things. This is a class about owning a wrench. Cynicism doesn't protect you, by the way; cynical people just get sold anti-hype instead. A ranking system protects you."

Say "I've bought things" like you mean it — have your own example ready if someone asks after class.

0:08 The three shelves

Goal: install the class vocabulary. Everything after hangs on these three words.

"I spent twenty-nine years behind a bar, so here's the bar version. Top shelf: SUPPORTED — multiple real studies, done by different people, pointing the same way. Middle shelf: SUGGESTED — one study, or a directional finding. Interesting. Worth watching. Not settled. And the well: SOLD — a vendor's own number, a testimonial, a vibe with a font. Say them with me as they land." (The words build one at a time — let the room say each.) "Nothing on the bottom shelf is necessarily false. It's just unverified — and priced like it's top shelf."

The hand gesture: high, middle, low. Use it every single time a claim gets shelved for the rest of the class — by the exercise, the room will be doing it with you.

0:12 The 60-second study autopsy

Goal: hand over the four questions that do all the work.

"Four questions, and you can run them in the checkout line. One: who was studied? Kids are not adults. College students are not you. And almost nobody studies ADHD adults over fifty — hold that thought, it gets its own slide. Two: how many? Fifteen people is a pilot, not proof. Three: was there a control group? If nobody compared the thing to not-doing-the-thing, you can't say it caused anything. Four: who paid? A company grading its own homework can be honest — but it starts on the bottom shelf and has to earn its way up."

Someone asking "what's a control group?" — never assume the room knows. The plain answer: "a group that didn't get the thing, so you can see what would've happened anyway."

0:16 Exhibit A: AI assists — it never treats

Goal: plant the top-shelf anchor and the instant-tell.

"First exhibit, top shelf. Twenty-seven clinical experts, peer-reviewed: AI should complement human judgment, never replace it. And here's what makes this finding useful rather than just humble — every credible product in this space says the same thing about itself. So you now own an instant tell: anyone promising that AI treats, cures, or fixes ADHD just told you which shelf they live on. The claim is the confession."

0:20 Exhibit B: fast capture is the real move

Goal: show what convergence looks like — the reason this one earns top shelf.

"Why is this one SUPPORTED? Because four roads that never met each other all arrive here. A peer-reviewed diary study caught an ADHD adult saying it better than any slide I could write: in 2 seconds I can scratch that itch, get it out, and go back to what I need to do. David Allen — the Getting Things Done guy — has said for decades: if capture isn't instant, it won't get done. Coaches use AI to beat blank-page freeze. And users keep reinventing brain-dumping on their own. When strangers keep converging on the same answer from different directions, that's what top shelf looks like."

0:24 Exhibit C: smaller beats bigger

Goal: the one rule, now with receipts.

"The same twenty-seven experts found AI's plans too generic — walls of text that re-overwhelm the exact brain they were meant to help. Economists found raw AI activity jumps forty to a hundred and eighty percent — but finished, shipped work rises only about thirty. Feel the gap between those numbers; that gap is where the hype lives. And grassroots users, no researchers involved, independently invented 'ask me one question at a time.' Put it together and you get the rule this whole program runs on: if AI doesn't make your next step smaller, it's not helping." (Lines reveal one per press — land the 30% before the rule.)

0:28 Exhibit D: the novelty wiring

Goal: model how to hold a SUGGESTED finding — believe it gently.

"Middle shelf now, and watch how we hold it differently. A 2018 brain-imaging study suggests ADHD brains weight novelty about twice as heavily as other brains — and that chasing the novelty predicted worse performance, not better. One good study. So I say 'suggests,' I don't say 'proves,' and if it fails to replicate someday, we'll take it off the wall. But notice how much it would explain: AI is novel, instant, and judgment-free — built out of exactly what this wiring craves. Every tool before AI asked that brain to suppress the craving. That's why AI feels different. And it's also why it can eat your afternoon — same circuit, both directions."

0:32 Exhibit E: hyperfocus is not flow

Goal: retire the superpower story without taking anything away from anyone.

"You've heard hyperfocus called the ADHD superpower. Measured head to head, flow and hyperfocus barely correlate — around r equals zero point one two, which in plain words means they're different animals wearing the same coat. Flow is voluntary; hyperfocus picks you. ADHD students report more hyperfocus and less flow than their peers. And no study shows hyperfocus produces better output than ordinary focus. I know that stings a little — the superpower story felt good. But you don't need a superpower to deserve tools that fit your brain. The wiring story is better anyway, because it's true."

Deflation. Follow the honest hit immediately with warmth — the class's whole brand is that respect sounds like accuracy.

0:36 Pop quiz: which number is real?

Goal: demonstrate how real findings get laundered into fake ones.

Read all three aloud, let the room vote by hands before anyone taps. The reveal: the MIT number is real — 453 professionals, writing tasks, general population. Then the teaching beat: "Option one is the same number wearing a costume — someone took a real general-population finding and quietly made it an ADHD claim. That's the most common trick you'll meet: real number, wrong population. The monitor number was never real anywhere."

0:40 The zero-studies club

Goal: the class's most shareable slide — famous things with no research behind them.

"Now the club nobody advertises. Body doubling — beloved by thousands, almost no formal research; we say 'people swear by it,' never 'proven.' Multi-monitor setups for ADHD — zero dedicated studies; that one's pure forum wisdom. Bullet journaling, morning pages, gratitude journaling — zero ADHD-specific controlled trials among them. Intermittent fasting for ADHD — zero studies, and we checked, so that's a confirmed hole, not an unsearched one. And AI speed gains for ADHD adults — zero; every number you've seen was borrowed from general studies. Membership in this club doesn't mean these things don't work. It means nobody knows yet — and anyone selling certainty about them is selling."

One line per press. Pause after "and we checked" — it's the credibility line of the night.

0:45 The most overlooked finding of all

Goal: land the 50+ gap with care — for many rooms this slide is personally about them.

"One more number, and this one's personal for a lot of rooms I teach. ADHD in adults over fifty is the most underdiagnosed age band there is — roughly zero point zero five percent have a recorded diagnosis, where studies expect three to five percent actually qualify. That's not a rounding error; that's generations of people who were called scattered, lazy, or 'a lot' — who just never got the word for it. So if you're in that band and this stuff is new to you: you're not late to this. Science is late to you." (Hold the silence after. Don't rush to the exercise.)

This is the slide that produces after-class conversations. Never diagnose, never speculate about anyone — the line is "that's a doctor conversation, and it's worth having."

0:50 Exercise: the claim autopsy (10 min)

Goal: the skill leaves the slide and enters their thumbs.

"Phones out. Find one ADHD claim in the wild — an ad, a TikTok, a product page, a supplement bottle if you've got one in your bag. Run the four questions: who was studied, how many, control group, who paid. Then give it a shelf. Write the claim and the shelf on your card. And the standing rule if you type anything into an AI while you dig: postcard rule — fake names, nothing you wouldn't write on a postcard." (Start the timer. Walk the room.)

People who found a claim but can't answer any autopsy question — that IS the finding. Tell them: "If you can't find who was studied in sixty seconds, that's your answer. Sold shelf."

1:02 What did you catch?

Goal: three real examples teach the ranking better than thirty slides.

Take three volunteers max. For each: have them read the claim, then ask the room — hand gesture ready — "which shelf?" before the volunteer gives their verdict. If verdicts disagree, celebrate it: "Good — shelving is a skill, not a formula. What question would settle it?"

1:06 Even the footnotes lie sometimes

Goal: the twist — citations themselves need the autopsy.

"Here's the part that surprised me most building this program. We fact-check every source behind these classes. Some citations pointed at pages that don't exist. One supplement claim cited a real paper — which, when we actually read it, found the opposite of what was claimed. And four numbers in our own earlier notes turned out subtly wrong; we caught them before they reached a slide. The moral isn't 'trust nothing.' The moral is: a footnote isn't proof, it's an address. Sometimes nobody's home. When a claim really matters to you — money, health — knock on the door."

"So how do I check a citation?" — the honest, practical answer: "Paste the claim and the source into your AI and ask 'does this source actually say this?' Then spot-check the quote. It's not perfect. It beats trusting the font."

1:10 About your For You page

Goal: give social media its fair, shelved verdict.

"The one peer-reviewed study of ADHD TikTok found exactly what you'd guess: huge reach, weak clinical accuracy. I'm not going to tell you to quit it — half of you found out ADHD-anything exists from that feed, and that has real value. The verdict is narrower: it's a lead generator, not a doctor. Let it point you at questions. Don't let it hand you answers. And when a video quotes a number — you own the wrench now."

1:14 Pop quiz: the autopsy reflex

Goal: confirm the reflex is installed, and preempt reflex-cynicism.

Room votes by hand first. The distractor worth naming on the reveal: option three, "assume it's false and move on." Say: "Cynicism feels smart and teaches you nothing. The four questions take sixty seconds and work on every claim you'll meet for the rest of your life. That's the trade."

1:18 What's left standing

Goal: resolve the tension — after all that debunking, what survives?

"After ninety minutes of taking claims apart, here's what's still standing when the dust settles. Capture thoughts the second you have them — supported from four directions. Make every next step smaller — supported by experts and economists who've never met. Put structure between impulse and action. Treat privacy as a rule, not a footnote. Notice anything? That's the Yap Method. Not because we built the class to flatter the method — the method was built FROM this microscope. It's what's left when you throw out everything that doesn't survive."

1:24 Become the evidence

Goal: turn the research gap into the homework.

"Remember the honest hole — nobody has studied us yet. So run your own study, n equals one. One week, three numbers, on paper: tiny actions actually finished, restarts from a save card, and minutes between deciding and starting. If the numbers move, you have evidence for the one brain that matters. If they don't move, you just saved yourself a subscription. Either way you win, and either way — you're doing better science than the ads are."

1:28 Trust is earned in shelves

Goal: close with the vocabulary in their mouths.

"Three words, one more time, together." (SUPPORTED · SUGGESTED · SOLD build one at a time — say them with the room. Confetti fires.) "Take the worksheet. Go be harder to sell to. And go easy on yourselves this week — that part's free."