The smart money says everyone will soon hold the same tools. The edge will belong to the minds that use them differently.
A few weeks ago, I sat on a panel at the Daily Maverick Marketing Masterclass, under a title designed to make marketers sweat: “Artificial Intelligence: Marketing’s Multiplier or its Executioner?”
The room was full of clever people holding confident opinions, which is to say it was a normal marketing event. At some point, I heard myself say a sentence I had neither fully planned nor thought through. “In the age of AI, the neurodivergent will inherit the earth.”
It landed. People wrote it down. Afterwards, I worried, as one does about a good line, that it might be too good to be true. So I did the thing I ought to have done before saying it in public. I went looking for whether the evidence agreed with me. But first, let me tell you how I got here, and why I said it.
The multiplier-or-executioner framing is, I think, the wrong fight. The more uncomfortable question, the one my fellow panellist Jarred Cinman kept circling, is this: when every brand can reach for the same generative models, the same predictive tools, the same data plumbing, where does an advantage come from?
Drowning in a sea of sameness
If every chief marketing officer is feeding near-identical prompts into the same handful of systems, the output will be technically competent and instantly forgettable. The much-feared sea of sameness is not produced by the machines. It is produced by people who mistake the tool for the thinker.
Which is why I reached for the neurodivergent mind: not the most fluent prompter, not the team with the largest AI budget, but the person who notices the connection nobody else saw and asks the question nobody else thought to ask. The tool is identical for everyone. The mind holding it is not. In an age of AI, humans are the differentiator, so surely neurodivergence will be a major benefit.
The theory sounds poetic, like it must be right, but the research I did afterwards showed me I was being sentimental. It was a humbling moment. But important because research and data are important to me. Sentiment can never trump insights.
Let’s examine the myth I had half-swallowed myself together: that attention-deficit minds are natural-born context-switchers, evolved for a working day of multiplying tabs and interrupting tools, and therefore quietly built for the AI office.
The myth of switchers with superpowers
It is an attractive idea. It is also, on the evidence, wrong. A review in Frontiers in Psychiatry found that adults with ADHD show higher error rates when switching between tasks, with those costs tied to working-memory limits, rather than the opposite of a switching superpower. Confuse novelty with fragmentation, and you will design a workplace with more interruptions, then congratulate yourself on being inclusive. You will have achieved the reverse.
The honest account is duller and more useful. Some people with ADHD do their finest work in fast, novel, interest-rich conditions, and their worst in routine, delay and administrative drag. That is not a story about chaos-loving brains. It is a story about the fit between a person and an environment, which is one of the few things leaders can actually shape.
But the research reveals that once you drop the cliché, real strengths come into view, and they are specific. The strongest signal sits at the front end of the enterprise. A recent meta-analysis from Syracuse University’s Whitman School, pooling dozens of studies and hundreds of effect sizes, found that the hyperactive and impulsive traits associated with ADHD were positively linked to entrepreneurial attitudes and behaviour, while inattention dragged on the outcomes that come after launch.
In plain terms: promising at the start, harder in the middle. That maps neatly onto where work is already drifting, towards smaller units, solo operators, fractional roles and one-person businesses, with software absorbing the back-office grind. The age of AI may suit the neurodivergent founder far better than the neurodivergent middle manager. They are different jobs and, often, different people.
A preference for the written word
For autistic professionals, the evidence points somewhere equally concrete. Studies tracking real-world behaviour, including work from Leiden, find a consistent preference for written, computer-mediated communication, such as email, messaging, and text, over spoken communication.
Almost everything about AI-assisted work is written, asynchronous and explicit. A government evaluation here in Britain of a Microsoft Copilot rollout found overall satisfaction at 72 per cent, but markedly higher among neurodiverse staff at 80 per cent, who were also keener to recommend it. One participant said the tool had “levelled the playing field.”
In a separate study presented at the 2024 CHI conference, autistic adults preferred advice from GPT-4 to a human stand-in for navigating awkward workplace exchanges. The point is not that the machine is wise. It is that an always-available, text-based intermediary reduces the strain of an interaction that weighs more heavily on some minds than on others.
None of this is news to the organisations that started early. SAP has run its Autism at Work programme since 2013 and reports retention well above ninety per cent. JPMorgan Chase, EY, and others have built similar pipelines and published striking productivity figures for particular technical roles.
Designing work well
I cite these numbers with one hand raised because they describe carefully designed programmes (with support, mentoring and matched roles) that do not leave people to fend for themselves. That caveat is the entire argument of the insight. Roughly fifteen to twenty per cent of us are neurodivergent, many without knowing it, so this is not a niche accommodation. It is a question about most workforces, including yours.
What do you think about this research? More importantly, if you are neurodivergent, I’d love to hear your thoughts. Do these insights sit well with you?
The writer Jurgen Appelo, who is autistic, offers the cleanest way to hold all this. He dislikes the word prosthesis for what AI does, and so do I. A prosthesis implies something is missing. Like a limb to be replaced, a deficit made whole by machinery.
The better word, he suggests, is adapter: the small device that lets one system speak to another without either being broken. Google Translate is not a prosthesis. The plug adapter in your suitcase is not a prosthesis.
Wired for success
Neither system is faulty; they are simply wired differently, and the adapter bridges them. My experience with my own neurodivergence and with neurodivergent colleagues is that this is precisely the potential AI represents. It is a transformational translation layer between one kind of mind and a working world built, until recently, around another.
Here is where I have to take my own bumper sticker apart. The romance of the “neurodivergent superpower” assumes that more stimulation, more mediation, more always-on synthesis are straightforwardly good. For minds that already run hot, it can become an amplifier. The first sign is usually higher output, which managers adore. The second is exhaustion. The third is burnout, which they tend to notice too late.
If you’ve read my piece ‘A brave transformation of shared value’, you’d see that overproduction is an industry norm we’ve shattered at the Brave Group. In many ways, the marketing sector is home to neurodivergence. At Brave, shared values, alongside deliberate investment in career-progression conversations and performance transparency, were introduced during one of the agency’s most commercially challenging periods. We held to them anyway.
An age of fluidity
The fashionable promise that neurodivergent people will flourish in fluid, ever-reteaming organisations does not survive contact with the research either: constantly changing membership erodes trust and shared understanding, while predictability, clear communication and explicit support are exactly what most autistic workers say they need. Fluidity is a coordination cost to be managed, not a gift for which everyone should be grateful.
So the useful shift is from labels to design. Stop asking whether ADHD or autism is an “advantage” in the abstract. Ask instead what a role actually demands. Enquire about its switching load, ambiguity, sensory and social intensity, and tolerance for error, and then build around that signature.
Cap concurrent priorities. Put handoffs in writing. Define what “done” means. Cut the surprise meetings. Let the software draft, recap and decompose, and keep human judgement where the cost of an error is legal, ethical or reputational. It is unglamorous work. It is also how work is genuinely repaired.
Adapting to rote machine logic
There is a deeper point for those of us who lead, and it is the one I keep returning to in the leadership rooms where I teach. For most of the industrial era, we asked human beings to adapt to the machine. By this, I mean aligning the schedule, the production line, and the standardised assessment.
The difference was filtered out long before it could contribute, which is part of why so many capable autistic adults remain unemployed or underemployed to this day. That is a loss of talent, not of ability. What is shifting now is the direction of accommodation. For the first time, the system can bend towards the mind rather than the mind being forced to bend towards the system.
That is the real content of my too-good line. The neurodivergent will not inherit the earth by birthright, as though a diagnosis were a winning ticket. They will inherit it wherever leaders are deliberate enough to build the conditions in which difference becomes contribution.
In my own writing, I drew a distinction between ladders, which you climb in a single prescribed direction, and trampolines, which let you launch from wherever you happen to land. The age of AI is a trampoline for the differently wired, but only if someone bothers to build the frame.
The future of work is not a gift. It is a design choice. As it happens, I can live with that. So, I suspect the rest of us can, provided we choose well.
