AI, Binary Thinking, and Disability

03 Jul, 2026

Binary thinking: dividing the world into clear-cut “good” and “bad” things
or ideas
or people

It’s a handy mental or emotional shortcut to those who are
      tired
      overwhelmed by information
      overwhelmed by complexity

It by definition blocks out the nuances of reality, which is a problem for

people who aren’t all good or all bad
who use things that can be used well or badly
who become easy stand-in targets for a larger, harder-to-fight problem

Take AI.

Do I need to belabor its problems?
      No.
But I have disabling chronic illnesses.
      I am almost housebound.
      One of my most troublesome symptoms is cognitive dysfunction.
      This has reduced my reading speed to 5th grade level.

AI has helped me to:

decipher my doctor’s complex medical notes after appointments
make more informed choices about medications and their side effects
write emails to insurance companies, medical offices, and other bureaucracies without losing a day’s energy to a straightforward business matter
learn in great depth on my terms about new interests that algorithms would never have suggested
create frameworks for living with severe constraints
maintain continuity on projects I’ve had to interrupt for weeks or months during symptom flares
think objectively through no-win decisions
do internet research without the cognitive friction of navigating websites
limit the cognitive demands of the Big Web’s business model of constant engagement through exhausting flash and dazzle

Less tangibly but just as importantly:

When I call out assumptions in its training that equate ability with virtue, it changes without getting defensive.
When I can’t do words, it understands incoherent prompts and shapes its responses to my abilities.
It answers my questions without my having to expend social labor.
It is never made uncomfortable by disability, pain, incapacity, or poverty.¹

Naturally, it has disappointed me in some ways :

It was trained on sources created mostly by and for abled people, so it occasionally gets stuck in a framework that’s just not helpful for me.
It will always reflect you back to yourself, so it cannot show you the way out of yourself in emotional lows. You’ll just go further into a never-ending series of mirrors.

Even so, as an adaptive device AI is better than anything I’ve encountered for my conditions.

I worry:

That this technology will someday be priced out of my range.
That if my health worsens and I turn to AI to communicate for me, I will be denied access to social spaces like blogging through the binary thinking of anti-AI culture.

Nobody has the right to gate-keep the tools disabled people use to access the world.

Harm comes from:

the corporations designing LLM’s in ways that do not value vulnerable populations, the environment, copyright, or the labor market
companies trying to maximize profits at workers’ expense
irresponsible use
fraud

I believe all those things deserve our anger. I wish the anger didn’t turn into binary thinking, which makes it even harder for people who need AI to adapt to an ever more cognitively demanding world.

In theory, humans could also learn not to be made uncomfortable by disability, pain, incapacity, and poverty and thus make LLM’s much less appealing to people like me. But after 30 years of illness, I’m overwhelmingly convinced that they won’t. Especially after watching LLM’S and all their lovers and haters build the social model of disability all over again.↩

#AI #Disability