TL;DR: Neurodivergent people spend roughly 80% of their cognitive lives translating their experience into a language designed for brains that don’t work like theirs. The kicker is that the very working memory fragility that creates complex needs is the same system required to articulate those needs in real time. The struggle to explain is the diagnosis itself. It’s not simply a communication failure.
I was taking notes during a fight with my wife.
Not to win. Not to build a case. I was trying to hold the shape of the argument in my head. When Charlotte and I used to go round-and-round (our anxious-avoidant attachment dynamic was one of the first things I discovered through journaling with AI), the conversation would branch and double back and loop, and my working memory couldn’t hold the map. I lose track of which problem we’re solving. I forget what was said three minutes ago. I can feel that there’s a structure to the conflict, but I can’t see all of it at once without writing it down.
So I was taking notes. Trying to build the map externally, because I couldn’t build it internally.
Charlotte didn’t see map-building. She saw me “going back to catch her in something.” Building ammunition. Lawyering up mid-argument.
“I was taking notes so I could respond to you in a safe and loving and respectful way,” I told her.
She heard: “I’m documenting this.”
Same act. Two completely different interpretations. I couldn’t explain the real reason in real time, because explaining it would require the very working memory capacity I was trying to compensate for. To make her understand why I needed the notes, I’d need to hold her emotional state, my own emotional state, the neuroscience of working memory fragility, our attachment dynamics, and a coherent argument all in my head simultaneously.
At one point during that fight I said something I didn’t have the framework to understand yet:
“I have a pattern of hearing things that you’re not saying. And I don’t know how to fix that. I don’t know how to fix that part, and I wish I did.”
I was describing working memory fragility. I just didn’t know that’s what I was going to call it yet. My brain was dropping context mid-conversation, backfilling with pattern-matched assumptions, and I couldn’t tell the difference between what she actually said and what my nervous system predicted she’d say. I was naming the problem in real time and had zero language for it.
I don’t actually remember what that argument was about. The working memory didn’t hold it. I only know these details because I recorded the conversation and later synthesized the transcript with AI as part of building my own cognitive scaffolding. The external system held what my brain couldn’t. Which is, of course, the entire point.
That’s the thing nobody tells you about being neurodivergent. The hardest part isn’t having different needs. It’s that explaining those needs requires the exact cognitive resources that are already impaired.
The Translation Tax
I call it the Translation Tax. Every time I need something from another person, the request itself requires me to run four parallel processes simultaneously:
First, I have to reverse-engineer my own need. Why do I need this specific thing? My body knows. My nervous system knows. But translating felt-sense into language means routing through working memory, which is the precise system that’s fragile in my architecture.
Second, I have to translate from constellation thinking into linear, sequential language. My brain doesn’t process reality in neat thesis statements. It processes in pattern-clouds, spatial relationships, felt resonances. Compressing that into “I need it quiet” is like describing a symphony by listing the notes in order.
Third, I have to manage the emotional labor of potential misunderstanding. My core wound is feeling unseen. Every explanation is a bid for connection that might land as “you’re being difficult” or “that’s a lot” or just a blank stare that tells me I’ve once again failed to translate my operating system into something legible.
Fourth, all of this is happening while my working memory is already running on fumes. Because that’s the whole point. The needs exist because working memory is fragile. And now I’m asking that same fragile system to construct a real-time argument for its own existence.
That’s four full-time cognitive jobs just to ask for a glass of water.
And this isn’t just intimate relationships. I spent fifteen years in product management, which turned out to be the perfect career for a brain like mine, because I could spend long nights translating my constellation thinking into detailed JIRA stories and strategy roadmaps. Asynchronous translation. My workaround. But every Friday demo told the same story: engineers had skimmed the specs, absorbed maybe 20% of what I’d painstakingly translated, and built accordingly. I did the translation perfectly. It still didn’t fully land. The tax isn’t just on the sender. It’s on the entire pipeline.
(That’s probably its own article. For now, back to the relationships where the stakes are higher than a missed sprint goal.)
The Architecture, Not the Attitude
Earlier this week I published a piece called ‘You’re So Smart!’ Is Not the Compliment You Think It Is. The comments were validating. One reader said that being praised for intelligence while struggling with basic tasks created a deep internal dissonance: “The praise raises your status in other people’s eyes, but internally it amplifies shame. Instead of feeling seen, you feel like a fraud.” Another said the “broadband intellectual throughput paired with dial-up executive function infrastructure” line was the reason she’d never believed she was actually smart. A third said she was “gobsmacked” reading someone describe thinking at the speed and complexity she does, because for years she thought she was alone.
Twelve likes, seven restacks, and a handful of people who clearly felt described for the first time.
This article is the companion piece to that one. Because the same architectural reality that makes “smart” a useless compliment is what makes explaining your needs feel impossible. It’s all the same root system.
My ADHD paper lays out the neuroscience: the ADHD brain isn’t a broken version of the neurotypical brain. It’s an alternative cognitive architecture. An Explorer phenotype that routes cognitive traffic through the basal ganglia and cerebellum rather than relying primarily on the prefrontal cortex. Procedural memory is fully intact (my rowing stroke is automatic, my fingers know the keyboard, I can drive a car without thinking about the mechanics). Working memory is the specific bottleneck (you should see me try to learn a new board game: holding the rules, tracking game state, planning two moves ahead while remembering what everyone else just did. It’s rough).
This is the dissociation that explains everything. Not a global deficit. A specific architectural trade-off.
And here’s what that trade-off means for communication: the system responsible for holding information in mind, manipulating it, and translating it into sequential language is the exact system that’s fragile. The system responsible for pattern recognition, deep processing, and felt-sense understanding is running at full power. I know what I need. I know it in my body, in my nervous system, in the spatial architecture of my mind. I just can’t export it into words fast enough.
It’s like having a high-resolution camera with a dial-up upload speed. The image is sharp. The bandwidth is the bottleneck.
The Five Layers
Let me map the actual terrain, because this isn’t one problem. It’s a recursive stack of five problems that compound on each other.
The Invisible Infrastructure Problem. Most of my actual needs are for external working memory support. Things that look “extra” or “unnecessary” to someone whose brain holds context internally. Try explaining that you need 300 browser tabs open or your entire thought architecture collapses. I’ll wait while they suggest bookmarks. (I’ve tried bookmarks. Bookmarks are where my thoughts go to die. I still have bookmark folders with “Netscape” in the name. One is labeled “From Phone - 4/17/10.” That’s fifteen years ago. I have never reopened it.)
The Preemptive Shame Shield. My nervous system has learned, across four decades of social data, that my needs get pattern-matched to “difficult,” “needy,” or “overthinking it.” So I’ve developed a protective mechanism: don’t ask. Don’t explain. Just absorb the friction and mask through it. The cost of not asking is high, but the cost of asking and being misunderstood is higher, because that cost hits my rejection wound directly. And rejection wounds don’t forget.
The Legitimacy Question. When your needs take ten minutes to explain, part of you starts interrogating yourself. “Am I making this complicated?” No. You’re operating with different base requirements. It’s the same as someone with diabetes explaining insulin timing. The need is real. The architecture is just different. But the self-doubt creeps in because the world keeps telling you that simple needs should have simple explanations, and yours never do.
The Stochastic Resonance Irony. Here’s a layer most people miss entirely. My brain requires a specific threshold of external stimulation to function optimally. Research shows that white noise improves ADHD cognitive performance to levels comparable to stimulant medication. What degrades my performance is the absence of the right kind of noise and the presence of the wrong kind. But try explaining to your family that you need the TV on at a specific volume while also needing the dog to stop licking himself in the same room because one is signal and the other is interference. The look you’ll get is the look I’ve gotten a thousand times. The look that says: “You’re being ridiculous.”
The Meta-Irony. The very working memory fragility that creates these complex needs also makes it harder to articulate them in the moment. I need external scaffolding to explain why I need external scaffolding, but I can’t build the explanation without the scaffolding already in place.
It’s recursion hell. And there’s no Stack Overflow for this.
The 3,000-Word Problem
I wrote a Substack Note recently about how different reader types experience my long-form articles. Think about this: I write 3,000-word deep dives about working memory fragility for an audience whose working memory is, by definition, fragile. I’m building context-rich explanations for people whose core challenge is holding context.
This is the paradox of the entire newsletter. And it’s the same paradox as the explanation problem, just zoomed out. Every neurodivergent person who reads this article is performing the exact cognitive labor I’m describing, in order to understand the description of that cognitive labor.
(If you’ve made it this far, by the way: your working memory is doing something remarkable right now. Notice that.)
The four reader types I identified break down like this. The neurotypical reader who finds this intellectually interesting but doesn’t feel it in their body. The ADHD reader who’s been skimming, hit a line that locked them in, and is now hyperfocused. The AuDHD reader who’s simultaneously cataloging every claim against their own experience and feeling a deep, quiet relief at being described accurately. And the gifted/2e reader who already sees where this argument is going and is three paragraphs ahead of me.
All four are valid. And none of them are reading the same article.
What “Neurotypical” Even Means (If Anything)
I’ve started questioning whether “neurotypical” is a coherent category at all. It might just describe people whose masking hasn’t been questioned yet. People whose cognitive trade-offs happen to align with the environment well enough that the trade-offs remain invisible.
Everyone’s working memory has limits. My Dad wanders the halls of every hotel we stay at during family vacations because he can’t remember the room number. That’s not ADHD. That’s just a brain being a brain. Everyone translates internal experience into imperfect language. The difference is degree and consequence. When your cognitive architecture fits the default assumptions of your culture, you don’t notice the translation. When it doesn’t, translation becomes a second full-time job.
Framing neurodivergent needs as “special” or “extra” implies there’s a normal baseline. But there isn’t. There’s just a set of assumptions baked into how we communicate, work, and relate, and some of us fall so far outside those assumptions that the gap becomes visible.
That visibility isn’t the problem. The assumption of a default is the problem.
So What Now
The brutal reality: I cannot fully explain my cognitive architecture to everyone who needs to understand it. Not in real time. Not without exhausting myself and triggering every rejection wound I’ve collected since kindergarten.
But here’s what I’m learning.
The Pre-Loaded Bridge. I write these articles during my best cognitive hours, when working memory is relatively stable and I can construct the explanations I can’t build on the fly. Then I send the link. “I wrote something that explains this better than I can right now.” The article becomes external scaffolding for the conversation.
Strategic Disclosure. Not everyone gets the full architecture. Charlotte gets the detailed version (and she’s earned it: she recently sat down with my neurocomplexity coach to understand me from his perspective, not just mine). My neurofeedback practitioner gets the technical version. My colleagues at my day job just know me as the guy with all the documentation. They have no idea what’s behind it. I have a handful of twice-exceptional women I meet for what I can only describe as sapiosexual coffee dates: four-hour conversations about depth, complexity, and the specific loneliness of processing reality at high resolution. Charlotte encourages these. She doesn’t want that conversation, and she’s honest about it, which is its own kind of love. I call this my distributed intimacy model: different people for different cognitive needs, because no single relationship can hold all of what this architecture requires. Your partner gets more depth than your dentist. Your 2e coffee dates get a different depth than your partner. That’s not dishonesty. It’s honest infrastructure.
The Documentation Habit. Pre-write explanations when you have the cognitive capacity. Reference them later. I wrote a document called “How Jon Works” and keep it in my digital back pocket. A user manual for my own brain, ready for anyone who wants it. Not everyone will read it. That’s its own filter. But the point is: I can’t build that explanation in real time during a conflict, but I can build it at 7 AM on a Tuesday when my working memory is fresh and my prefrontal cortex is cooperating. This is what my Life Model system does at scale: holds the context so I don’t have to rebuild it from scratch every time I need to explain myself to a new doctor, therapist, or the person sitting across the table wondering why I’m scribbling during a fight.
The Enough Protocol. Sometimes I just state the need without explanation. “I need to write this down. It’s a brain thing.” People who deserve to be in your inner circle will accept this. People who need a twenty-minute justification for every accommodation are telling you something about the relational safety they’re offering. Listen to that.
For the middle ground (friends who want to understand but don’t need the full architecture), I keep pre-written texts in my digital back pocket. When someone sends me something I can’t respond to in the moment, I fire off one of these:
“My process is recursive rather than linear. I integrate inputs when my mind cycles back to that part of the build. So even if you don’t see me act on this right away, know that it’s in the queue and will resurface when it’s optimal in my workflow.”
Or the shorter version:
“I work in pattern loops, so the follow-up might happen hours, days, or weeks later depending on where this piece fits in the larger structure.”
I don’t care how crazy they sound. I wrote ‘em once, during a Tuesday morning when my prefrontal cortex was cooperating. Now I deploy them without rebuilding the translation from scratch each time. Which is, again, the entire thesis of this article in miniature.
The Struggle Is the Evidence
Here’s what I want you to sit with.
The fact that you can’t easily explain your needs is not a failure of communication. It’s the architecture itself, showing up in real time. The struggle to articulate is the most honest evidence that the need is real. You don’t struggle to explain things your brain handles effortlessly. You struggle to explain things that require the very system that’s impaired.
This is what Working Memory Fragility looks like in the wild. Not as a theory in a paper. Not as a diagnosis code on a form. As a person scribbling notes during an argument because their brain can’t hold the shape of the conversation without them, while the person they love most reads that scribbling as hostility. And then not even remembering what the fight was about. Only knowing it happened because the external scaffolding held what the brain couldn’t.
If you’ve spent your life feeling like your needs are too complex, too much, too hard to justify: they’re not. Your architecture is simply different. And the world hasn’t caught up to that yet.
But it will. Because people like us keep writing these damn articles, even when the writing itself requires the exact thing we’re writing about.
And if you just read 3,000 words about the impossibility of holding context, and you’re still here, holding context? That’s not a small thing. That’s your architecture doing something remarkable in service of understanding itself.
You don’t need to explain that to anyone.
Human. Deeply seen.
Jon Mick is the founder of AIs & Shine, building cognitive scaffolding for minds that work differently. He spent roughly 80% of the cognitive energy available for writing this article on translating what he already knew into words you could read. The other 20% went to remembering where he saved the draft.