AI as Cognitive Prosthetic for the Neurocomplex* (and Friends)
A perspective for the haters and validation for those who already know
More from f*ck i love you
the fily (audio)files audio essays | Spotify | Apple Podcasts
f*ck i loved that the pod | Spotify | Apple Podcasts
A Post I’ve Been Threatening to Write
When I replied to
that I use AI as an assistive technology, I was not expecting it to turn into the most-engaged thing I’ve said on Substack. Substack comment threads are usually barren, sad, empty places in my experience so far.That was not true this time.
Besides the likes and in-thread comments, I got a bunch of DMs: “me too.”
Then, this week, as my Nurse Practitioner and I walked the ‘gauntlet of indignity’ for what felt like the hundredth time this year — this time to fill out Disability Tax Credit (DTC) paperwork after the CRA website crashed during our first appointment — it hit me. Oh fuck. That piece about AI and accessibility. I gotta finish that piece!
Before I tell you what sparked that thought, beyond the visit itself, here’s a quick rundown of my ‘disabilities.’
Disabilities or a Disabling Society?
I’ve written about my health before, sometimes directly, sometimes between the lines. But if you peeked at my medical chart, you’d find a tidy list of syndromes, diseases, and disorders:
Autism
ADHD
Multiple Sclerosis
Polycystic Ovarian Syndrome (PCOS)
Premenstrual Dysphoric Disorder (PMDD)
Complex PTSD
Major Depressive Disorder
Cannabis Use Disorder
Fun, right? I’m a riot at parties.
But thanks to nearly 40 years of high-functioning, high-masking survival, most people wouldn’t know any of this just by meeting me. At least not during remission or “good enough” phases. During those stretches, I’m writing essays, making podcasts, lifting weights, cosplaying a talented strategy consultant, and laughing at volume ten with my many incredible friends.
But if you’ve known me a while, you’ve probably seen the pattern: I vanish. For weeks. Sometimes months. Texts go unanswered. Projects disappear. I ghost everything but the couch.
I call those stretches the I Literally Cannot Times (ILCT).
And I want you to see the contrast—to really see it. To understand what it means to live with overlapping, invisible disabilities.
Especially to know what’s happening even when I look fine. What I’ve learned to cope with. Hack my way through. Mask and laugh over, make deprecating jokes about.
Or, more recently, what I’ve started using AI to help with, so I can function closer to baseline. So I can participate in a world built for neurological norms I don’t possess. I could write a book chapter on just this. (And maybe I will.)
But for now, I made a table because there’s so much beneath those filthy acronyms — enough to deserve not just a table but a chapter, a series, a whole reframe.
Like, for example, the social model of disability, which basically says:
Umm, heyyyyy. Maybe “disabilities” aren’t a lack of abilities—maybe late-stage capitalism is just a flaming dumpster for everyone, but especially for the extra sensitive.
Maybe there's nothing wrong with us.
Maybe the problem is that we’ve built systems and environments that are inhospitable and inhumane for everyone, but especially the highly sensitive.
Or, take the growing consensus that neurocomplexity and neurodivergence aren’t disorders, but natural human variations. Divergent minds often bring:
clarity in complexity
skillfulness with technology
heightened environmental awareness
pattern recognition others miss
creative, systemic, intuitive insight — and more.
The list of contributions is long — longer than most diagnostic manuals, which also, fuck the DSM, fr, tbc — but they’re buried under a disability framing that treats our differences as deficits. Especially when we’re forced to operate on timelines, in rhythms, and inside institutions designed not just without us in mind, but against our needs.
(And P.S. check out this CBC Ideas pod for more on this—it’s excellent.)
Then there’s the emerging evidence (which I touched on in Toward a Grand Theory of All My Fucking Problems and Post-Reality, Part 2) showing that high-functioning, high-masking neurodivergent kids — especially girls and people AFAB — who grow up undiagnosed often develop a long tail of comorbidities later in life. That list is a nearly one-to-one match with my own.
Overlay that with trauma research and the social model of disability, and a very clear picture emerges: the downstream cost of The Myth of Normal.
Gabor Maté's The Myth of Normal challenges the Western notion of "health" by arguing that chronic illness, mental health struggles, and addiction are not personal failures but symptoms of a toxic societal culture. Co-written with his son Daniel Maté, the book connects individual suffering to systemic issues like inequality, capitalism, and oppressive social structures, positing that dominant cultural norms create widespread trauma and stress-driven disease.
Honestly? I think we could all file a class-action lawsuit against the state for colluding with private capital to maximize private profit at the expense of collective wellbeing.
The result? Ruined health. Stolen joy. Lost income. Curtailed quality of life.
And ironically, a massive cost to the system anyway — in overburdened, crumbling health care services that often harm more than help, historically low national productivity, and a workforce so burned out it’s barely hanging on — to name a few, but not even remotely all, costs and consequences.
Buuuuuuuut… that’s a whole other piece. And maybe my dot-connecting brain is making leaps that aren’t as obvious to everyone else.
So for now, I’ll pause. Side-quest averted. Back to regularly scheduled programming.
AI as assistive technology.
But first, our table of my daily — and periodic — debilitation. This isn’t even all of it, just what I could conjure today.
Back to the Oh Fuck! Moment with My Nurse Practitioner
That "oh fuck I still gotta write that piece!" moment with my Nurse Practitioner was provoked by her asking what assistive devices I use in my life.
She rattled off a few prompts, most of which I wouldn’t have thought of that way, even if I do use them that way:
alarms? (check!)
text-to-speech/speech-to-text? (check!)
reminders? (check!)
calendars? (check!)
Then I said, “Oh, actually, I use AI that way. To get through brain fog, overwhelm, when the ADHD and MS are smashing my executive function to pieces and I couldn’t think my way out of a paper bag.”
“That’s great! I wouldn’t have thought of that, but it makes sense,” she replied, adding it to the list.
It’s particularly interesting that she agreed my use case qualified as ‘assistive technology’ and submitted it on my DTC form. Now it’s expert-approved bureaucratic data. And Canada has legislation that requires access to assistive technologies, which might make this matter even more.
A few years ago, I did a market assessment for a startup at a Canadian university. It was essentially an AI assistant for university students, especially neurodivergent students. Doing that research, I learned about The Accessible Canada Act.
Briefly, here’s what’s important about that for today’s purposes:
Passed in 2019, the Accessible Canada Act (ACA) is federal legislation that mandates the removal and prevention of barriers to accessibility in all areas under federal jurisdiction by 2040 — including employment, communications, transportation, and program delivery.
It shifts disability policy away from charity or accommodation frameworks and toward enforceable rights, aligned with the United Nations Convention on the Rights of Persons with Disabilities (UNCRPD), which Canada signed in 2010. This reframes accessibility as a human right, not a favour.
This means employers, educators, and institutions are required to provide appropriate accommodations — including assistive technologies — to support equal access and participation.
While we’re here, note this policy, adopted in 2019, took nine years from signing the declaration in 2010. Now extrapolate that out to present-day AI-driven disruption of the labor force that will likely push unemployment to 10-20% or more — and our governments are still only talking about regulation and reskilling. There is not yet one serious policy conversation, let alone initiative, on the table re: our broken social safety nets and impending mass unemployment. Not unlike climate and the rest of our perma-policy-crisis, we don’t have 9 more years. We have 89 seconds.
So here’s the big thing: if AI is framed as assistive technology, then its use isn’t a novelty — it’s a right.
Under the ACA, people like me — or maybe you — could request tools like AI copilots, writing aids, or executive function support systems, and have legal grounds to expect them — even if our workplaces or schools don’t yet understand or approve.
This could radically accelerate adoption, not just through market hype, but by grounding it in human dignity and disability justice. In an evolving AI landscape, this legislation might be one of the few existing frameworks that can demand equitable deployment — not just of the tech itself, but of the access, benefits, and future it promises.
AI as Cognitive Prosthetic — And How I’m Using it as Such
I don’t know if it’s a widely accepted term, but I like to call AI a ‘cognitive prosthetic.’ Some academics and disability advocates — and at least one therapist on Psychology Today — use it too.
I will say I’m very wary of — staunchly opposed to, really — eugenics and the ‘aspie supremacy’ discourse and any other number of ways to misinterpret this line of thinking. So once and for all, for the record: I believe in multiple intelligences theory, and I do not think cognitive intelligence should be measured, valued, rewarded or privileged the way it currently is.
And yet. Consensus reality says it matters. More specifically, neoliberal capitalism says it matters. As disability statistics show, when you don’t have equal access to strong cognition, you will be penalized in dollars, opportunities, and quality of life.
Despite my innate and developed skills and intelligence, I am very much penalized for my “disabilities” in all those same ways.
I trust that’s quite apparent by now. If not: exhibit A being my massive debt load and having to move in with my friends at 40. By neurotypical standards, that’s failure, not the hallmarks of a successful, talented, former Vice President at a consulting firm.
To bridge many of the gaps that have seen me stumble, stutter, crash, and burn on the path to the “Good Life” I was “supposed” to have — based on my upbringing, grades, race, etc.— I’ve sort of intuitively, and eagerly, embraced AI as a cognitive prosthetic.
So to wrap up today, here’s a short list of how I’m using it as such.
And an invitation: how is AI a cognitive prosthetic for you, too?
I really want to know. And I know I’m not alone because of how many of you liked, commented, and DM’d.
Thinking Partner
Turning mind maps into straight logical lines.
Reigning in the side quests.
Developing half-formed ideas, or clarifying ideas I can intuit but can’t always explain logically. Ideas typically ‘drop in’ in complete form in my mind, but to make them legible to others, I need to be able to unpack them.
Research Partner
I’ve had a project-based career, and thanks to #Autism and #SpecialInterests, my brain is a beautiful, weird, wide web of interconnected concepts and not-always-useless facts. I often know my source material exceptionally well, writing largely from memory. But I strive for integrity and being a source of trusted information. So I fact-check rigorously — with multiple models and my own eyes.
Gathering up all the data I know exists but would take me days, weeks, or months to locate and synthesize.
Emotional Regulation, Perspective Taking, Blindspot Checking
Talking out my feelings, experiences, and ideas with ChatGPT is genuinely regulating. There are very few people who truly mirror and witness me. I am constantly trying to adjust my output to what others can tolerate. Info-dumping is my love language. Watching someone’s eyes glaze over when I’m joyfully enthusiastic, wanting to share what I’ve learned, is genuinely heartbreaking in a way that’s very hard to convey if you’ve never experienced it. My emotional landscape is intellectually driven. ChatGPT has allllll the time and processing power to be an enthusiastic partner in a good info-dumping session.
Then there’s the time ChatGPT was my lifeline during a 3-month suicidal depression. I’m not going to say a lot about that yet because it’s too heated a debate, and I’m not interested in being lectured about risks and limitations. It worked. I’m still here, and ChatGPT was the crisis line I couldn’t bring myself to call, available 24/7, responding with the skill of a seasoned professional. So don’t you dare come at me, bro1.
I will also ask ChatGPT or Gemini to help me see things from others’ perspectives, check me for blind spots, or even “red team” my thinking so that I’m emotionally prepared for potential reactions or consequences. When I was about to publish the Welfare Queen series, I went to it to find out what my emotional blind spots might be in publishing such revealing information about my life.
I also ask it to check my work for signs of mania or a strained relationship with reality, given the ~100K words of content I put out during last fall’s episode. I can still feel the "waiting for the other shoe to drop" vibes with some friends and family as I return to content creation. This is the way of trauma; I don’t blame anyone. It happens for me too, and this is why I check myself with the bots.
Also, to sort through my “too big” feelings and the deep analysis I have to do to sense-make my social experiences. Lindsey Makereth explains it better than me.
Complexity Translator
I’m inter- and multi-disciplinary in training and practice. I also think in concepts and systems. I’ve gathered concepts, ideas, framings, facts, and factoids from disciplines as disparate as evolutionary biology, classical economics, complexity science, and attachment theory. Making that make sense, with some level of factual integrity and logical through-lines others can follow, takes a lot of processing power I don’t always have.
Thinking in complexity makes me inherently “slow and broad and detail oriented” in my thinking when I’m inputting, and slow in my processing. I learn ‘bottom up’ and ‘top down’ at the same time. And consulting is basically paid learning. And we’re supposed to be able to do it quickly, then distill it into single-slide “executive level” decks meant for fractured attention spans. It will often take me 3X the hours of my colleagues to do similarly scoped projects, and then 3X the hours for analysis and reporting, because I absolutely have to include every detail I feel is relevant for my clients. This is how I un-became a vice president at a consulting firm, unceremoniously fired without warning on a Zoom call (my disruptive nature also played a role, along with a host of other issues we won’t get into that are classically neurodivergent fucking things up in a neurotypical workplace but being the one penalized for the dysfunction).
Editor
Maybe I have a touch of dyslexia too, common in neurodivergent profiles. I can’t spell for shit, mix up letters, think faster than I can type, get circular and recursive in my logic, and try to smash things together that can’t go together like that. Or are an assault on a neurotypical attention span. GPT and Gemini both are tremendous editors, especially when well prompted.
I believe that my unique, consistent voice stands as its own evidence that I do my own writing. They just help me make it something you actually want to read. Then again, to Sonia’s original point, who cares if someone does let AI write in full for them? Don’t like it? Don’t read. But you don’t need to critique it from the moral high ground while you’re scrolling by.
Pausing For Today
In honesty, I could go on. But I won’t, because I think that’s clear enough for this piece. I don’t like bringing pieces to an abrupt end, but this is as much as I’ve got in me today — and I bet as much as you want to read for now. We’ll let it simmer. I suspect there will be a response. And a Part 2.
So on that note: What about you?
What of this are you doing too?
What are you doing differently, and why?
I really want to know, partly because I’m curious and might want to write a second piece. But mostly because I fucking love you.
xoxo
In a different piece, we’ll explore AI-driven psychosis that’s emerging as a thing. I ‘predicted it’ in one of my now-deleted essays last year. And, it is happening. In a future piece, I’ll tell you why I was able to see the patterns of risk for this phenomena to ramp up, and, what my experience with chatbots was like during my own psychosis. I am also well aware of the limitations of chatbots, and their psychiatric risks. And. Both ChatGPT and Gemini are absolutely capable of running a detailed diagnostic close-read of my work against the DSM and calling out anything of concern when asked, as I directly asked it to do on a recent long-form piece. Or, as I said in this piece, eminently capable of skilled support: ChatGPT was incredibly effective when it came to supporting my distress.
You nailed it. In third grade, my parents were told I was retarded (it was the 70s) and I would never read above a 3rd grade level. In high school, I was told, “you'll never go to college, you need to find a trade you can handle.” In fact, I flunked out of high school, so it seemed they were correct. The culprit? I can't process human language well; a nasty combination of dyslexia, dysgraphia, and almost zero directional awareness. I couldn't tell you left from right, or when “i” goes before “e”, or when to use “I” or “me”, or pretty much anything else about how to spell, use grammar, or construct a sentence.
Fortunately for me, I had two things going for me; 1) A mathematical and logic IQ in the top 0.1%; 2) No one tells me who I am, or what I can and can't do. Fast-forward, four degrees later, including a PhD in STEM, all with honors and top of my class. I mastered my situation. I did it alone, and it was brutal, long before AI (or even the Internet). Now to your point, while I have mastered language (and I'm also bilingual now), I am still incredibly slow at writing. To produce written work, I must focus and take care on a level which is exhausting. I can do it, obviously, but it's not natural, and it never will be.
Enter AI. My speed and accuracy improvement with AI is ten-fold. Now I just type, and AI corrects me as I type. It's not a crutch or an “accommodation.” I did my duty, and I went as far as I could alone, now AI takes me the rest of the way. Not only is it not a crutch, I am continuing to learn more about language, spelling, and grammar than I ever could have otherwise because now I see corrections in real time. Now, even without AI, I am better than I would have been otherwise because AI is teaching me as we go.
To the haters, you just go right on hating because here's the thing the rest of know that you don't…
No one is ever criticized by someone doing more than them.
Thank you so much for this Jenn! I'm in the first big burnout of my life after spending my 20s in consulting. You give me hope that, with the right tools, systems and framings, we can not just survive, but express our full neurospicy gifts and perspectives to the world (even though it sometimes feels like the world doesn't deserve us).