%20(1).jpeg)
The first time I asked an AI model to help me draft a follow-up email to a client, I gave it the same shorthand I'd give a coworker who already knew the account: "Can you draft the follow-up to confirm we're still on track?" What came back was polite and generic. It didn't know which project, which deadline, or what "on track" even meant in this case. So I tried again, spelling out what that coworker would already understand: "Confirm we're on track for Friday, and flag that the scope note still needs their sign-off." That one landed. First try, exactly what I needed.
Most people write this off as a quirk of the technology. It's worth pausing on, though, because the fix is simple to state: be direct, be specific, don't assume shared or unspoken context, say exactly what you mean. That's advice neurodivergent communities have been giving the rest of the world for years. AI isn't creating a new communication style. It's forcing mainstream adoption of one that already existed.
A 2024 YouGov poll found that 19% of U.S. adults, nearly one in five, identify as neurodivergent. Chances are you already work with someone who is, and you don't even know it. "Neurodivergent" describes brains that process information and communication differently from what's considered typical. It covers autism, ADHD, auDHD, dyslexia, dyspraxia, and more. It's a wide range, because it affects every person differently. The core idea is that these are natural differences, not deficits to fix.
Different processing doesn't mean less capable. Neurodivergent people bring exceptional depth to their work: intense focus, sharp pattern recognition, and a level of detail and rigor in problem-solving that's genuinely hard to find. The same wiring that makes small talk exhausting can also make someone the person who catches the error everyone else missed.
You may not know you work with a neurodivergent person because of a common coping strategy known as masking: when a person minimizes outward signs of their differences and reactions to sensory overload to read as more "typical." Masking takes extreme cognitive effort. It's like hiding your true personality, and it makes it that much harder to focus on the things that actually matter. Without enough recovery time, it can lead to burnout. It also takes more energy to infer what typical conversation leaves unsaid: the subtext behind "can you draft the follow-up," the assumption that context will fill in the gaps. Ambiguity itself is often the hard part, not the task. Direct, explicit communication is just better, not only for neurodivergent minds, but for everyone.
That's the connection to AI. It has the exact same dependency on stated context that neurodivergent communities have pointed to for years. Working with a chatbot makes that pattern visible, and gives everyone else a reason to recognize it in the people already sitting next to them.
When neurodivergent and neurotypical people miscommunicate, it isn't because one side is at fault. It's a two-way mismatch in expectations and context, and both sides contribute to it.
AI gives neurotypical people a fast, low-stakes way to feel that mismatch themselves. A vague prompt produces a technically correct but wrong answer, and the instinct is to blame the tool. But that instinct, treating a literal interpretation as a failure rather than a legitimate response to an ambiguous request, is exactly what neurodivergent people have described encountering in conversation for years. The tool doesn't lack empathy. It lacks the shared context a vague prompt assumes. That's a useful, humbling parallel.
There's a well-known idea in accessibility circles called the curb-cut effect: features built for people with disabilities, like curb cuts, closed captions, and voice interfaces, end up benefiting everyone. Something similar is happening with AI. Clear agendas instead of an open-ended chat. One question at a time instead of three buried in a paragraph. Explicit success criteria instead of implied ones. These are the exact accommodations neurodivergent employees have long requested in meetings and workplace communication, and often dismissed as "extra" needs.
Turns out they also produce better AI output. When a practice framed as an accommodation for some people quietly becomes the standard that makes a tool work well for everyone, it's worth asking why it took a chatbot to make the case.
Making your workplace more inclusive doesn't require knowing who on your team is neurodivergent, and that's the point. You don't need a diagnosis on file to build a workplace that doesn't force people to mask just to function. The curb-cut effect holds here too: clearer agendas, plainer instructions, and more flexible formats help the neurodivergent employee who's been struggling quietly, and everyone else who was too polite to admit the vague meeting invite was useless.
Build in structure by default. Put instructions, assignments, and feedback in writing instead of relying on something said once out loud. Make priorities explicit when multiple tasks land at once, instead of leaving people to guess what matters most. Give clear deadlines and expectations, and advance notice on schedule or meeting changes instead of same-day surprises.
Give people some control over their environment: a quiet, private workspace or a place to retreat when things get overwhelming, a personal light instead of overhead fluorescents, a desk away from high foot traffic, flexibility on dress code when certain fabrics are genuinely painful to wear, and remote or hybrid options instead of a fixed in-office schedule.
Make room for people to communicate the way they work best. Remove the unnecessary cognitive load, and people do better work. That's the bottom line for the business, not just the individual. Inclusion isn't a separate initiative bolted onto how a team already operates. Done well, it's just better design, and AI just gave everyone a reason to notice and put better practices into play.
I still think about that client email. The version I got when I relied on unspoken context, and the version I got when I was clear and direct. The gap between those two wasn't the tool getting smarter. It was me consciously improving my communication, the way the neurodivergent people around me had been asking me to for years.
