I read the cover twice; glanced through the interior a few times before turning my attention to my own thoughts. The message underneath all that clinical language was short, and it was nothing I hadn’t already been saying to myself for years.

Something is wrong with you. It needs fixing.

Maybe you have your own version of it, in an office or in your own head at 2 a.m., after the third missed deadline or the friendship that went quiet and never came back. The verdict lands the same. You are the problem. You are the thing to fix.

I won’t tell you the struggle isn’t real. The missed deadlines were real. The avoided emails, the shame, the flat exhaustion of running a brain to which the world keeps handing the wrong instructions – all real. I’m not here to turn your hard days into a gift.

But “broken” was never the right word.

It’s the grinding that happens when one kind of brain meets a world that was quietly built for a different one. That is a design problem. Not a character flaw.

And knowing that gives the knower a certain kind of power over it.

Where the fix points

There are two ways to describe the exact same neurodivergent life, and they are far from the same.

The first is the medical model, and it sees a list of symptoms to treat, manage, or shrink. In that story the friction lives inside you, so the repair does too. You are the broken thing. Try harder, mask better, buy the planner, get easier for the room to hold.

The second is the social model, and it sees a mismatch instead. A brain meets a room that was quietly built for somebody else. In that story the friction lives in the gap between the two. And once you can see the gap, you get to ask a different question. Not what is wrong with me, but what is incompatible within the environment.

When a wheelchair user can’t get into a building, we don’t send the wheelchair back for a quick-fix. We look at the stairs. Nobody stands in the lobby wondering what’s wrong with the person for needing a ramp.

Most of our mismatches are harder to see than a flight of stairs. So we went on believing that we were broken wheelchairs.

We weren’t.

The open-plan office that makes focus impossible is a design choice. “Just send a quick email,” said like it’s one step and not eleven, is a design choice. The calendar, the form, the rule nobody wrote down – all of it was built for a brain that isn’t yours, then handed to you like the trouble was your fault.

Naming that is not a word game. It’s the moment you consider where to put the ramps in the room. This is what we talk about here. It is called The Cognitive Ramp for a reason, after all.

I want to return to the idea of the instruction manuals from a moment ago. I have often been handed the wrong instructions to life. But when I received my AuDHD diagnosis, I felt like I’d been handed the web address to my own personal instruction manual. All I needed to do then was decode it.

This issue’s prompt: two ways to say the same thing

It’s one thing to understand the difference between the medical model and social model. It’s another to feel it. Here’s how to start doing that.

Take one thing you’ve long considered one of your broken wheelchairs. The task you always avoid. The way you lose an hour. The thing you’re “bad at.” Then paste this into Claude, ChatGPT, or whatever AI assistant you use:

I want to describe one thing I struggle with, two different ways. Here’s what actually happens: [describe it plainly – what you’re trying to do, and where it breaks down]. First, describe it in medical-model language, as a symptom I need to manage. Then describe it again in social-model language, as a mismatch between how my brain works and how this task or environment is set up. Keep both short. Then tell me: which version points at something I could actually change?

Read them side by side. One will tell you to fix yourself. The other will show you where the ramp goes.

Here’s my own example:

I want to describe one thing I struggle with, two different ways. Here’s what actually happens: I repeatedly make bad financial decisions, such as buying a computer case to fit a second GPU when my motherboard doesn’t support one (a fact I overlooked at first). First, describe it in medical-model language, as a symptom I need to manage. Then describe it again in social-model language, as a mismatch between how my brain works and how this task or environment is set up. Keep both short. Then tell me: which version points at something I could actually change?

And here is Claude’s response:

Medical-model version (symptom to manage):
You have a pattern of impulsive, poorly-checked spending — a deficit in impulse control and planning. The buying gets ahead of the verifying. It’s a symptom of how your attention and executive function work, something to catch and suppress before it costs you money. The fix lives inside you: try harder to slow down, resist the urge, remember to double-check.

Social-model version (mismatch):
Buying hardware requires holding several dependencies in mind at once (does the motherboard support a second GPU? does the case fit? does the PSU?) and checking them in a specific order before committing. The purchase flow doesn’t enforce that order — nothing stops you at checkout to confirm compatibility. Your brain runs ahead to the goal; the task assumes you’ll manually gate it. The gap between those two is where the money leaks out.

Which one you can actually change:
The social-model version. The medical version tells you to be a different person — more restrained, more vigilant — which is roughly the thing that isn’t reliably available to you, so it mostly generates guilt. The social-model version points at the environment, and environments are editable: build the compatibility check before the buy button instead of relying on memory. A pre-purchase checklist for anything over $X. PCPartPicker, which refuses to let incompatible parts into the same build. A 24-hour hold on hardware orders. A rule that you paste the spec into a compatibility checker before paying.

None of those require you to have better impulse control. They move the checking out of your head and into a system, so the ordering happens whether or not you remember to do it. That’s the difference the two framings expose: one asks you to fix your brain, the other asks you to fix the checkout.

The word was never broken. Neither were you. The room just needs a door where somebody assumed a wall.

If something loosened a little, subscribe – and forward this to one person who’s spent too long assuming they were the problem.

Subscribe now

Share

– Brian

This newsletter sits deliberately between those two extremes.

It starts from a premise that shapes everything I write: neurodivergent brains are not broken. ADHD, autism, AuDHD, dyslexia, OCD, and other forms of neurodivergence are valid ways a brain can be organized. The struggle so many of us live with is not proof of personal failure; it is friction between a cognitive style and environments that were designed around different assumptions. When a wheelchair user cannot enter a building, we do not blame the wheelchair. We build a ramp.

That is where this newsletter gets its name.

I believe modern AI — the kind of conversational assistant you can talk to in plain language — can be one of those ramps. Not a cure. Not a replacement for therapy, medication, community, or accommodations. A ramp: a piece of cognitive accessibility technology that can hold steps outside your head, translate vague expectations into concrete actions, draft the email you have been avoiding for three days, and give you a patient, nonjudgmental place to think out loud at two in the morning.

That possibility is new. For most of history, this kind of support required another person — a body double, an assistant, an unusually understanding friend — and access to those people has never been distributed fairly. Today, for the first time, a flexible thinking partner is available to almost anyone, at almost any hour, for little or no money, with no technical skill required. If you can send a text message, you can do everything I will ever describe here.

Who I am

I’m Brian Amaro-Jeppesen — a writer and AI practitioner with AuDHD, OCD, and anxiety. For those who don’t know, AuDHD is sort of an intersection where Autism and ADHD, two conditions that are in many ways opposites, meet in all sorts of tricky ways. I spent years asking “What is wrong with me?” before trading it for a better question: What does my brain need to thrive, and how can these new tools help me build it? I’ve spent the last few years building AI-powered support systems for my own brain, testing them in my own days, refining them through my own setbacks, and keeping only what actually worked. I wrote a book about it — AI For Neurodiversity, coming soon — and this newsletter is where those ideas live between now and then.

What you’ll get

Every two weeks: one idea and one prompt. The idea will be short — a single concept about how neurodivergent minds meet a world built for different assumptions, and what to do about it. The prompt will be something you can copy, paste, and use the same day. No hype, no doom, no jargon, no homework. Skim it, skip issues, come back whenever — there is no wrong way to read a newsletter about there being no wrong way to have a brain.

This issue’s prompt: the two-minute first step

For when you know exactly what needs to be done and cannot start. Paste this into Claude, ChatGPT, or whatever assistant you use, and fill in the blank:

I need to do [task], and I can’t start. Don’t give me a pep talk, and don’t explain why starting is hard — I know. Break this down and give me only the smallest possible first physical action, something I can do in under two minutes. Then stop and wait. When I tell you it’s done, give me the next step. One step at a time only.

The magic is not in the AI. The magic is that the steps now live outside your head, the next action is always exactly one, and nobody is sighing at you while you get there.

If this sounds like your brain — or like someone you love — subscribe, and forward this to one person who needs a ramp. See you in two weeks.

— Brian