"There are basically two ways to know you have a future," Palantir CEO Alex Karp told an audience in March 2026. "One, you have some vocational training. Or two, you're neurodivergent." The statement, delivered with the casual authority of a man whose company is valued at over $200 billion, travelled quickly through every professional network that trades in career anxiety. Within days, Palantir posted a dedicated Neurodivergent Fellowship, describing neurodivergent individuals as people who "will play a disproportionate role in shaping the future of America and the West" and who "see past performative ideologies and perceive beauty in the world that still exists."
Karp is hardly alone in this enthusiasm. Gartner predicts that by 2027, one-fifth of sales organisations within Fortune 500 companies will actively recruit neurodivergent talent to improve business performance. Microsoft has published extensive case studies of senior executives with ADHD and dyslexia who describe AI tools as transformative for their daily work. Fast Company ran a feature titled "AI is levelling the playing field for neurodivergent talent." The narrative has acquired a momentum that makes questioning it feel, at best, ungrateful.
Meanwhile, Anthropic co-founder Daniela Amodei offered a rather different account of what the AI era would demand. "The things that make us human will become much more important instead of much less important," she told ABC News. "We look for people who are great communicators, who have excellent EQ and people skills, who are kind and compassionate and curious." It is worth noting that these are precisely the competencies that many neurodivergent individuals find most effortful to perform, at least within the conventions of a typical corporate environment.
The tension between these two positions is instructive. They cannot both be entirely right, and neither is entirely wrong. What they reveal, taken together, is that the conversation about neurodivergence and AI has moved well beyond the research base that might reasonably support it, and that the stakes for getting the framing right are considerably higher than the current discourse tends to acknowledge.
The conversation about neurodivergence and AI has moved well beyond the research base that might reasonably support it, and the stakes for getting the framing right are considerably higher than the current discourse tends to acknowledge.
The New Survivorship Narrative
Karp's framing is significant because it shifts the status of neurodivergence from a condition requiring accommodation to a competitive advantage requiring recruitment. In this account, cognitive difference is not something organisations should tolerate or support. It is something they should seek out, because the AI-transformed economy will render conventional cognitive profiles insufficient. The neurodivergent mind, in this telling, possesses a natural immunity to the automation that will displace everyone else.
This is a seductive proposition, and it is not difficult to see why it resonates. For individuals who have spent years navigating systems that were not designed for their cognitive architecture, being told that the future belongs to them carries a particular emotional weight. For organisations that have struggled to articulate a business case for neuroinclusion beyond compliance, it provides a compelling commercial rationale.
The difficulty is that the claim rests on a series of assumptions that are, at best, partially supported. The idea that neurodivergent individuals are inherently more creative, more innovative, or more resilient to automation conflates a set of genuinely diverse cognitive profiles into a single category of advantage. ADHD, autism, dyslexia and dyspraxia are not interchangeable conditions, and the cognitive profiles associated with each differ in ways that matter profoundly for how work is experienced and how value is created.
ADHD, autism, dyslexia and dyspraxia are not interchangeable conditions, and the cognitive profiles associated with each differ in ways that matter profoundly for how work is experienced and how value is created.
What the Research Actually Shows
The relationship between neurodivergence and creativity is more conditional than the current enthusiasm tends to suggest. A widely cited body of research has associated ADHD with higher scores on certain measures of divergent thinking, particularly those that reward fluency and novelty. Individuals with ADHD appear, in some experimental settings, to generate a greater number of unusual associations than matched controls.
However, as a comprehensive review in Neuroscience and Biobehavioral Reviews has shown, these findings are far from uniform. The effect depends heavily on the type of creativity being measured, the context in which it is assessed and the presence of co-occurring conditions. A 2025 study from the University of Bath found that enhanced creativity previously attributed to autism was in fact driven by co-occurring ADHD, raising fundamental questions about how these traits are categorised and interpreted.
What is rarely mentioned in the superpower narrative is that the same cognitive profile associated with higher divergent thinking scores is also associated with difficulties in convergent thinking, the capacity to evaluate, select and refine ideas. In organisational settings, divergent thinking without convergent follow-through tends to produce ideas that remain unrealised, which is a pattern that can be as costly as it is interesting.
Divergent thinking without convergent follow-through tends to produce ideas that remain unrealised, which is a pattern that can be as costly as it is interesting.
Where the Real Shift Is Happening
The more productive question is not whether neurodivergent people possess innate superpowers, but how the introduction of AI into everyday work alters the conditions under which different cognitive profiles operate. The evidence here is more concrete and, in important respects, more encouraging than the mythology.
A 2025 study commissioned by the UK Department for Business and Trade found that neurodivergent workers were 25 per cent more satisfied with AI assistants than their neurotypical counterparts and were significantly more likely to recommend these tools. This finding aligns with the broader argument of this series. When AI reduces the cost of organising, structuring and presenting thought, individuals whose primary constraint has been organisational rather than intellectual stand to benefit most directly.
Microsoft's own research, published in partnership with EY, documented how employees with disabilities and neurodivergence reported significantly greater productivity gains from Microsoft 365 Copilot than their neurotypical and non-disabled colleagues. Kim Akers, a Microsoft executive living with ADHD, dyslexia and dysgraphia, described AI as having fundamentally changed her capacity to manage the administrative demands of her role, allowing her to spend less energy on translation and more on the work itself.
For individuals with ADHD, AI can provide the convergent scaffolding that their cognitive profile does not naturally prioritise. The capacity to generate ideas remains, whilst the capacity to structure, evaluate and present those ideas is augmented. For autistic individuals, AI can mediate the translation between deep analytical thinking and the social-linguistic conventions of professional communication, reducing the energy spent on code-switching rather than on substantive contribution.
When AI reduces the cost of organising, structuring and presenting thought, individuals whose primary constraint has been organisational rather than intellectual stand to benefit most directly.
What Neurotypical Workers Actually Face
The counterpart to this discussion, which tends to receive rather less attention, concerns the position of neurotypical workers in the same shifting landscape. If AI reduces the cost of precisely those tasks that neurotypical individuals have traditionally performed with relative ease, the competitive advantage that once came from being well-organised, articulate and procedurally reliable may diminish in economic value.
This does not mean that neurotypical workers will become disadvantaged. It means that the gap between cognitive profiles, in terms of observable output quality, is likely to narrow. A well-structured email, a competently organised presentation, a clearly formatted report, these outputs once signalled a specific kind of professional competence. When AI can produce them for anyone, the signal becomes less informative and the underlying thinking becomes more visible.
The EY Global Neuroinclusion at Work Study, published in 2025, found that many neurodivergent employees scored highly on leadership competencies and innovative thinking. At the same time, the study noted that nearly half of these individuals had not disclosed their condition at work, suggesting that the organisational conditions required to realise these advantages are often absent. The advantage, in other words, is not an inherent property of the condition. It is a function of the interaction between a cognitive profile and the environment in which it operates.
When AI can produce a polished email or presentation for anyone, the signal becomes less informative and the underlying thinking becomes more visible.
The Persistent Cost of Cognitive Friction
What the superpower narrative consistently underplays is the friction that neurodivergent individuals continue to experience in organisational environments designed around neurotypical assumptions. Meetings structured around real-time verbal processing, performance systems that reward consistent output over variable intensity, communication norms that equate brevity with competence and elaboration with inefficiency. These are not minor inconveniences. They represent a continuous tax on participation that AI can reduce but cannot eliminate.
Research published by Understood.org in partnership with the American Association of Advertising Agencies in 2025 found that nearly one in two people working in the creative industry identified as neurodivergent, yet a significant proportion reported that their workplace did not provide adequate support. The creative industries, which might be expected to value cognitive diversity most highly, appear to benefit from neurodivergent participation without systematically accommodating the conditions under which it occurs.
The cost of this friction is borne disproportionately by the individual, and it compounds over time. Burnout, underperformance relative to potential and career stagnation are not failures of talent. They are consequences of sustained misalignment between cognitive architecture and organisational design. Karp's fellowship may recruit neurodivergent candidates, but the question of whether Palantir's working environment is designed to sustain them over a full career is one that a recruitment announcement cannot answer.
Burnout, underperformance relative to potential and career stagnation are not failures of talent. They are consequences of sustained misalignment between cognitive architecture and organisational design.
The Risk of a New Hierarchy
There is a risk that the current discourse, rather than enabling a more honest engagement with cognitive diversity, simply replaces one hierarchy with another. If the old hierarchy favoured neurotypical competence profiles, the new one may favour a romanticised version of neurodivergent creativity. Both versions overstate what is known and underestimate what remains difficult.
Describing neurodivergence as a superpower creates an expectation that can itself become a burden. It implies that the value of cognitive difference lies in exceptional performance, which means that neurodivergent individuals who do not display visible creative brilliance are left without a narrative that legitimises their experience. The person with ADHD who struggles with deadlines, loses track of communication threads and finds open-plan offices genuinely debilitating is not well served by a framework that tells them their condition should make them the last professional standing in an AI-transformed economy.
AI tools can reduce the cost of certain cognitive operations, and for neurodivergent individuals, this reduction is often more significant than for those whose cognitive profiles already align with organisational norms. This is a meaningful shift, and it deserves serious attention from anyone responsible for how work is designed and evaluated. What it does not do is render the underlying challenges invisible, nor does it eliminate the need for structural accommodation in how roles are scoped, how performance is measured and how careers are sustained.
If the old hierarchy favoured neurotypical competence profiles, the new one may favour a romanticised version of neurodivergent creativity. Both versions overstate what is known and underestimate what remains difficult.
Beyond Mythology, Towards Specificity
Neurodivergent minds are neither uniformly disadvantaged nor uniformly gifted. They operate differently, and the consequences of that difference depend on what is being asked of them, in what context, and with what support. What changes with AI is not the nature of cognitive difference itself, but the economic and organisational conditions under which it is evaluated.
A more productive framing would move beyond both the deficit model and the superpower model towards something that acknowledges specificity, conditionality and the role of environment in determining whether a given cognitive profile produces advantage or friction. Karp is right that the AI economy will reward people who think differently. He is less right about who those people are, and he is not right at all that a neurological diagnosis is sufficient to predict who will thrive.
That kind of honesty is more difficult to market than a fellowship announcement. It is, however, rather more likely to produce organisations in which cognitive diversity is genuinely supported rather than merely celebrated. The structural reasons why such support so rarely materialises, even when the intention is present, are examined in the cognitive meritocracy illusion, which traces how internal evaluation continues to reward the appearance of capability long after the signals on which that appearance rests have lost their informational content. The full integrative argument, including the documented evidence on dedicated programmes at SAP, Microsoft, JPMorgan, EY, DXC and Auticon and the strategic decisions facing executives, is developed in the capstone essay of this series on the quiet reordering of talent.
Questions this essay answers
Is neurodivergence a competitive advantage in itself?
Neurodivergence is not, on its own, a competitive advantage. It becomes a source of competitive advantage when the firm has organisational conditions that admit non-standard cognition into the contribution flow, which in most firms requires deliberate redesign rather than rhetoric.
Where does the survivorship bias enter the public debate?
The most visible public claims about neurodivergent talent rest disproportionately on individuals who have already navigated organisational structures successfully. The population of neurodivergent professionals whose contribution was filtered out before becoming visible does not appear in the same statistics, which biases the debate toward the optimistic end.
What does the most recent evidence suggest?
The evidence from the 2024 Microsoft and EY Copilot study, the 2025 Gartner work on AI in the workplace and the UK Department for Business and Trade research points to material productivity effects from generative systems, with the largest absolute gains for employees who previously paid the highest signal-production cost.
Sources
- Karp, A., public remarks and Palantir Neurodivergent Fellowship announcement, March 2026.
- Amodei, D., interview with ABC News on AI and human skills, 2026.
- Gartner, "By 2027, 20% of Fortune 500 Sales Organizations Will Actively Recruit Neurodivergent Talent," 2025 forecast.
- Microsoft and EY, Neurodivergent Employees and Microsoft 365 Copilot: Productivity Impact Study, 2025.
- University of Bath, "Enhanced Creativity in Autism Driven by Co-Occurring ADHD," 2025.
- UK Department for Business and Trade, AI Tools and Neurodivergent Worker Satisfaction, 2025.
- Boot, N. et al., "Creativity in ADHD: A Review," Neuroscience and Biobehavioral Reviews, 2020.
- EY, Global Neuroinclusion at Work Study, 2025.
- Understood.org and American Association of Advertising Agencies, Neurodivergence in the Creative Industry, 2025.
About this Series
This article is part of the Synthetic Minds at Work series exploring how generative AI is reshaping the role of cognition in organisations.
