My family is neurodivergent. ADHD, autism, or both - depending on which one of us you ask. Most of my colleagues are too. At some point you stop calling it a coincidence and start calling it a pipeline.
The tech industry didn’t just accidentally attract neurodivergent minds. It was built by them. And the agentic age is about to make that relationship even more interesting.
We’re Everywhere
Stack Overflow’s 2022 developer survey found 10.57% of ~70,000 developers self-report ADHD. The general adult population rate is around 6%. That makes developers roughly 1.5-2x more likely to have ADHD than the average person.
The autism numbers tell a similar story. Simon Baron-Cohen’s research at Cambridge found twice the rate of autism in Eindhoven (a Dutch tech hub) compared to other Dutch cities. His team tested 500,000+ people and found STEM workers score significantly higher on autistic traits. 34% of young adults with ASD pursue STEM majors, versus 23% of the general population.
— Bill Gates, Source Code (2025)If I were growing up today, I probably would be diagnosed on the autism spectrum.
29% of entrepreneurs have ADHD. Adults with ADHD are 6x more likely to start their own business. This isn’t a coincidence. It’s self-selection at scale.
The Different Operating System
The “superpower” framing gets thrown around a lot. It’s partly right and mostly incomplete.
What’s real: ADHD brains are wired for novelty, pattern recognition, and lateral problem-solving. Coding triggers hyperfocus because it provides constant feedback loops and problem-solving dopamine hits. Autistic minds bring systematic thinking, logical consistency, and deep focus. Research shows autistic engineers produce more robust code with fewer errors.
What’s also real: Executive dysfunction is a career killer. A 2024 study found the ADHD-to-burnout pipeline is mediated by executive function deficits. Hyperfocus is fickle and activates on the wrong thing more often than the right one. Up to 80% of autistic workers experience burnout. 85% of college-educated autistic people are unemployed or underemployed.
Many neurodivergent people in tech spend enormous energy “passing” as neurotypical. Scripting conversations, forcing eye contact, suppressing stims, mirroring social norms. One developer described it as “acting all day, followed by collapse at home.” Remote work reduced masking energy by 30% in one study, but mandatory video calls clawed much of that back.
The honest framing isn’t “superpower” or “disability.” It’s a different operating system running on the same hardware. The environment determines which applications run beautifully and which crash constantly.
Tech Culture: The Accidental Fit
Tech didn’t set out to be neurodivergent-friendly. It just happened to build exactly the right conditions:
- Stimulating, novel problems: ADHD brains crave constant change. Tech rarely delivers boredom.
- Flexible hours: Working around medication timing, energy cycles, and hyperfocus windows changes everything.
- Async communication: Code reviews are text-based, precise, and on your own schedule. Precision over social performance.
- Output over presence: The best engineering cultures evaluate what you ship, not how confidently you spoke in a meeting.
- Remote work: Control over your sensory environment. No open-plan hell. No fluorescent lights.
I wrote about this from a different angle in my introverts post - many of the same shifts that help introverts thrive (remote work, async, documentation-first culture) are exactly what neurodivergent developers need too. The overlap isn’t surprising. There’s a reason these traits cluster.
HPE found neurodiverse teams were 30% more productive. SAP’s Autism at Work programme generated innovations worth an estimated $40M in savings. This isn’t charity. It’s competitive advantage.
The Orchestrator Era
Here’s where it gets interesting. And where I think nobody has connected the dots yet.
The traditional career path for engineers goes: IC, senior IC, then either stay technical or move into management. For neurodivergent developers, that management path is often brutal. You go from deep technical work with clear rules and immediate feedback to managing people with social dynamics, ambiguous communication, and endless meetings without agendas.
Many neurodivergent developers either stay as ICs forever or burn out in management within two years. I’ve watched it happen dozens of times.
The agentic age creates a third path: IC to orchestrator.
Addy Osmani’s framework describes the future developer as moving from “conductor” (directing one agent synchronously) to “orchestrator” (defining tasks for a fleet of agents in parallel). Human effort is front-loaded on specs and architecture, back-loaded on review and validation.
Think about what the orchestrator role actually requires:
- Designing systems of interacting agents: Pure systems thinking. Autistic strength.
- Writing precise specs and constraints: Explicit, rules-based communication. Autistic strength.
- Monitoring multiple parallel workstreams: Rapid context switching. ADHD strength.
- Experimenting with new tools and approaches: Novelty-seeking. ADHD strength.
- Deep architectural thinking when it matters: Hyperfocus on hard problems. Both.
- Comfort with non-linear, chaotic workflows: Agents produce results out of order. ADHD natural habitat.
No unwritten social rules. No office politics. No ambiguous feedback. You get leverage and scope through clear, explicit interfaces - without the social complexity that burns neurodivergent developers out.
AI as Executive Function
If ADHD is fundamentally a disorder of executive functioning, then AI is an executive function prosthetic. Not a cure. A scaffold.
- Context maintenance: The AI holds the codebase in memory so you don’t have to. External RAM for brains that leak memory.
- Task chunking: AI breaks overwhelming work into micro-steps. Task initiation is the hardest part for ADHD brains, and AI handles it.
- Documentation and boilerplate: The tasks that drain neurodivergent attention the most are exactly the tasks AI handles best.
- Capturing fleeting ideas: Chat interfaces catch thoughts before they evaporate from working memory.
— Yulia Rafailova, executive function coachAI offers an incredible tool to externalize our thinking, acting as a scaffold that can significantly reduce the cognitive load on our working memory.
EY found 85% of neurodivergent employees say generative AI creates a more inclusive workplace. A UK government study found neurodiverse workers were 25% more satisfied with AI assistants than neurotypical respondents.
The Other Side of the Coin
The tech-focused narrative makes neurodivergence sound manageable. Hyperfocus on code, ship features, dopamine loop keeps you going. But neurodivergence doesn’t clock out when you close your laptop.
- Brilliant at work, falling apart at home. You can architect a distributed system but forget to pay a bill or eat lunch. Executive function deficits hit everything that lacks a dopamine reward.
- Hyperfocus as avoidance. Coding for 14 hours straight isn’t always a superpower. Sometimes it’s dodging the difficult conversation or the life maintenance piling up for weeks.
- The dopamine cliff. Shipping feels amazing. Maintenance feels like wading through concrete. This is why neurodivergent developers leave a trail of 80%-finished projects.
- The “gifted but…” pattern. High-performing at work. Laundry piling up. Relationships strained. Health neglected. The gap between professional capability and personal functioning is the least discussed aspect of neurodivergence in tech.
And AI doesn’t fix all of it. It doesn’t fix burnout from masking. The 85% underemployment rate for college-educated autistic people isn’t because they haven’t found the right AI tool. The privilege filter is real: getting diagnosed, getting into tech, having access to AI tools - each step filters out people who need support most.
The point isn’t that ADHD makes you a better developer. It’s that the tools and environments are finally catching up to how different brains actually work.
The Next Generation
My kids are growing up in a world I didn’t have. They’ll have AI assistants that scaffold their executive function from childhood. They’ll enter a job market where directing AI agents is a core skill, and where the traits that made school difficult map directly to the orchestrator role.
I grew up masking, compensating, and burning out before I understood why. My kids will grow up with language for it and tools that meet them halfway.
That’s not a superpower. It’s just a better fit.
Related Posts
