Cognitive Wheelchairs
Why Anti-AI Virtue Signaling Marginalizes Invisible Disabilities
The public discourse surrounding generative Artificial Intelligence (AI) frequently reduces the technology to sweeping ethical condemnations. However, this conversation routinely ignores a massive demographic: the chronically ill and neurodivergent populations who utilize Large Language Models (LLMs) as vital assistive technology. For individuals navigating the complex, exhausting intersection of hypermobile Ehlers-Danlos Syndrome (hEDS) and Attention-Deficit/Hyperactivity Disorder (ADHD), LLMs are not a shortcut or a lazy gimmick—they are a literal cognitive wheelchair.
The Invisible Link: hEDS and ADHD Executive Dysfunction
To understand why LLMs serve as an essential prosthetic, one must first look at the medical reality linking systemic physical conditions to neurodevelopmental ones. Clinical research demonstrates a profound transdiagnostic connection between heritable connective tissue variants like hEDS and neurodevelopmental conditions such as ADHD (Sharp et al. 805). When a person has hEDS, genetic defects in systemic collagen do not simply cause joint dislocations and chronic pain; they radically alter the autonomic nervous system, body-mapping awareness, and baseline cognitive processing (Bulbena et al. 237).
This biological reality forces the brain to run in overdrive just to handle basic physiological stability. Impaired proprioception (the body’s awareness of its position in space) means maintaining basic motor competence can completely overload a person’s remaining executive functions (Kindgren et al. 380).
When you layer ADHD on top of hEDS-driven physical fatigue, the resulting executive dysfunction is severe. Simple administrative workflows, organizing linear thoughts, and overcoming task initiation paralysis become monumental, agonizing hurdles.
This is precisely where an LLM functions as a critical accessibility device. For a person dealing with a wheelchair for people with ADHD caused by hEDS, an LLM acts as an externalized cognitive scaffold. It handles the exhausting heavy lifting of initial text organization, breaks monolithic tasks down into bite-sized steps, and mitigates the crushing working-memory deficits that would otherwise cause a total cognitive system crash (Deshpande).
The Cognitive Prosthetic: Just as a physical wheelchair provides structural mobility when a person’s physical joints cannot reliably bear their own weight, an LLM provides structural scaffolding when the brain’s executive pathways are locked by systemic fatigue and neurodivergence.
The Elitist Backlash: Performative Outrage and Real Harm
Despite the clear utility of these tools for accessibility, the cultural conversation is currently dominated by an aggressive, exclusionary backlash. Frankly, it is complete horseshit to virtue signal against AI without examining who you are systematically shutting out of the digital economy in the process.
Online commentators, creative purists, and tech cynics routinely take to social media to comisserate and get attention for marginilizing a group of people unknowingly. By framing the use of any text-generative tool as an inherent moral failure, they convert a highly effective neurodivergent coping strategy into a cheap punchline for internet clout.
When critics sweeping compiler-level decrees that using an LLM to outline an essay, structure a professional email, or parse dense jargon makes someone a “lazy writer” or an “unoriginal thinker,” they are speaking from a position of intense neurotypical and able-bodied privilege. They treat the baseline capability to sit down, focus perfectly, and linearly translate chaotic thoughts into pristine syntax as a moral virtue—rather than what it actually is: a biological luxury.
The Wheelchair Double Standard
The hypocrisy of this performative outrage becomes blindingly obvious when you look at how these exact same critics interact with technology in their everyday lives. Let’s be completely real: most people have probably tried it. They have messed around with LLMs out of curiosity or convenience to draft a quick response, clean up code, or summarize an article. They fully understand how much friction the tool removes. Yet, they turn around and shame disabled individuals who rely on that exact same friction reduction just to survive in professional and academic spaces.
As a society, we have at least achieved a baseline level of visible empathy where people wouldn’t make fun of a physically disabled person for using a wheelchair. If a person lacks the musculoskeletal integrity to walk up a steep flight of stairs, we do not stand at the top of the steps mockingly telling them they are “cheating” the architectural design. We build a ramp.
Yet, when it comes to neurodevelopmental or systemic physical conditions that lack an obvious external marker, that basic human empathy completely evaporates into the ether.
The Accessibility Double Standard
Visible / Physical Accommodations (e.g., Paralysis, Joint Mobility)
Accepted Supports: Wheelchairs, ramps, automated lifts.
Social Perception: Universally protected; mocking someone’s physical aid is widely recognized as unacceptable and is heavily socially condemned.
Invisible / Cognitive Accommodations (e.g., hEDS-linked ADHD, Executive Dysfunction)
Accepted Supports: LLM text scaffolding, prompt assistance, AI parsing.
Social Perception: Actively shamed; routinely dismissed by critics as intellectual laziness, fraud, or a lack of originality.
Shaming someone with hEDS-induced ADHD for using an LLM to organize their thoughts is functionally identical to kicking a crutch out from under a patient’s arm and telling them to just try harder to stand. So what in the fuck makes them think its OK to make fun of invisible disabilities?
Performative internet activism shouldn’t come at the cost of human accessibility. It is time to end the short-sighted moral panic and acknowledge that cognitive accessibility tools deserve the exact same cultural respect, protection, and dignity as physical ones.
Works Cited
Bulbena, Antonio, et al. “Psychiatric and psychological aspects in the Ehlers–Danlos syndromes.” American Journal of Medical Genetics Part C: Seminars in Medical Genetics, vol. 175, no. 2, 2017, pp. 237-45. https://doi.org/10.1002/ajmg.c.31544
Deshpande, R. “The Confluence of Code and Cognition: An Analysis of Generative AI’s Impact on ADHD Diagnosis Trends Among High School Students in Northern California, 2022-2025.” Preprints.org, 2025.
"Explain how LLM's are like a wheelchair for people with ADHD caused by hEDS" prompt. Gemini, 2026 version, Google, 30 May 2026, gemini.google.com.
Kindgren, Erik, Antonia Quiñones Perez, and Rajna Knez. “Prevalence of ADHD and Autism Spectrum Disorder in Children with Hypermobility Spectrum Disorders or Hypermobile Ehlers-Danlos Syndrome: A Retrospective Study.” Neuropsychiatric Disease and Treatment, vol. 17, 2021, pp. 379-88. https://doi.org/10.2147/ndt.s290494
Sharp, Harriet Emma Clare, Hugo D. Critchley, and Jessica A. Eccles. “Connecting brain and body: Transdiagnostic relevance of connective tissue variants to neuropsychiatric symptom expression.” World Journal of Psychiatry, vol. 11, no. 10, 2021, pp. 805-20. https://doi.org/10.5498/wjp.v11.i10.805