The Playbook Is Dead

What Alex Karp gets right about neurodivergence, AI, and the end of the standardised mind

John Pope · Founder, midagent AI · Ottawa, Canada · April 2026

I have a personal stake in the argument I am about to make. I will not spell out all the details of my story here, but I will say this: I have spent much of my adult life being one of the most useful people in the room, while also being the most difficult to employ. The delta between those two points was never fully understood or explained to me — until recently.

Alex Karp, the mercurial, philosopher-king CEO of Palantir, has offered a cogent and vivid explanation about neurodivergence. And I think he is right — about this topic. Others topics not so much. Generally, he grates on me. But not this time.

In early 2026, speaking at the TBPN podcast and then at the World Economic Forum in Davos, Karp made a claim that cut through the usual noise about AI and the future of work. "There are basically two ways to know you have a future," he said. "One, you have some vocational training. Or two, you're neurodivergent." He followed it by launching Palantir's Neurodivergent Fellowship — a deliberate, unapologetic recruitment programme targeting people whose brains are wired differently. Not as a diversity gesture. As a competitive strategy.

The reaction was predictable. Some praised it. Many dismissed it as eccentric provocation from a famously eccentric executive. I want to argue that the dismissal of his beliefs about neurodivergent productivity is wrong from a fiduciary perspective — and that Karp is pointing at something most institutions are not yet ready to face.

What the Industrial Economy Was Actually Optimising For

To understand why Karp is right, you first have to understand what the last hundred years of economic organisation were actually built to produce. The answer is not intelligence. It is not creativity. It is compliance.

The industrial economy needed reliable, repeatable human performance. The assembly line required workers who would show up on time and perform the same operation the same way, without deviation. The corporate office required managers who would follow the approved process, navigate hierarchy with appropriate deference, and execute the playbook. The professions — law, finance, medicine, consulting — were built around the ability to ingest an enormous body of codified knowledge and apply it correctly within a structured framework.

Schools were designed to serve this economy. They optimised for retention and recall, for sitting still, for following multi-step instructions in sequence, for converting the full complexity of a student's mind into a single ranked number that employers could compare. The children who could perform this task well were rewarded. The children who could not were corrected — or diagnosed, or quietly sidelined, or told in a hundred small ways that the way their brain worked was a problem to be managed.

A neurotypical brain — wired for pattern compliance, social attunement, sequential processing, and rule-following — was genuinely the most useful instrument for most of the available work. The market valued standardisation. It built its institutions to produce standardisation. And it got what it paid for.

Then We Built the Machine

Artificial intelligence is, at its core, the ultimate realisation of everything the industrial economy valued in human beings. It follows rules with perfect fidelity. It ingests enormous bodies of codified knowledge and applies them at scale. It executes the playbook — legal discovery, financial modelling, structured code, routine correspondence, standard analysis — with breathtaking speed and zero fatigue. The entire edifice of reliable, predictable, repeatable performance that the economy spent a century building into human beings has now been automated.

This is the insight at the core of Karp's argument, and it is devastating in its logic. The people most exposed to AI displacement are not, by and large, those who were told their whole lives that they were difficult. They are those who were told their whole lives that they were doing everything right. The student who mastered the playbook. The analyst who produced flawless, standardised outputs. The professional who excelled at processing established frameworks. These people did not fail the system. The system trained them for a role that a machine can now perform better, faster, and cheaper.

"The people most exposed to AI displacement are not those who were told their whole lives they were difficult. They are those who were told they were doing everything right."

https://youtu.be/VTjsDNq1Eqw?si=QQWBnCEnycRucHDe

The rules are being re-written in the value of human labour.

The World Economic Forum's Future of Jobs Report 2025 projects that AI and automation will displace 92 million jobs globally by 2030, with 86 percent of businesses expecting these technologies to fundamentally transform their operations. The disruption is concentrated precisely in structured white-collar work — the roles that high-compliance minds were educated and credentialled to fill.

The Neurodivergent Advantage Is Not a Metaphor

When Karp talks about neurodivergence, he is not speaking abstractly. He is describing specific cognitive traits — the kind that come with ADHD, autism spectrum conditions, dyslexia, and related profiles — that are structurally resistant to the kind of automation AI currently does well.

People with ADHD frequently exhibit what researchers call divergent thinking — the capacity to generate multiple, non-obvious solutions to a single problem, rather than converging on the single correct answer the playbook demands. Many autistic individuals demonstrate what Cambridge psychologist Simon Baron-Cohen terms a high systemising quotient — an innate drive to identify deep structural patterns in complex rule-based systems, the kind of insight that underlies breakthrough engineering and novel architecture. Dyslexic minds frequently show exceptional visual-spatial reasoning and interconnected thinking — the ability to see relationships between seemingly unrelated concepts that a more linearly-pruned brain might not notice.

These are not soft advantages. They are the specific cognitive operations that AI currently performs worst: out-of-distribution reasoning, genuine novelty generation, pattern recognition in conditions of high ambiguity, and the willingness to discard a failing framework and build a new one from scratch. The neurodivergent mind's characteristic resistance to the playbook — the very trait that made it inconvenient in the industrial economy — is precisely what makes it hard to replace.

Karp credits his own dyslexia for Palantir's success, describing it as the thing that prevented him from mastering the conventional academic path and forced him to develop a more non-linear way of thinking. The research on dyslexia and entrepreneurship is striking: while roughly 10 percent of the general population has dyslexia, studies suggest that as many as 35 to 40 percent of entrepreneurs do. Something about the cognitive profile that makes standard academic performance harder also appears to drive the pattern recognition, risk tolerance, and unconventional problem-solving that building something genuinely new requires.

The Roll Call Was Always There

Karp is not announcing a new phenomenon. He is naming one that was hiding in plain sight.

Elon Musk publicly acknowledged his Asperger's diagnosis in 2021, linking the same brain that made social interaction difficult to the obsessive focus that made electric vehicles and reusable rockets possible. Richard Branson has dyslexia and ADHD; he dropped out of school at fifteen and has been outspoken about dyslexia being a different thinking skill-set, not a disadvantage. Steve Jobs is widely considered to have had dyslexia. IKEA founder Ingvar Kamprad had dyslexia and ADHD. JetBlue founder David Neeleman has ADHD and has described it as the source of his ability to think in ways that others cannot.

Venture investor Peter Thiel has argued that mild Asperger's can free entrepreneurs from what he calls an attachment to social conventions — the conformity pressure that prevents most people from building companies that look genuinely strange at inception. The greatest technology companies in the world did not begin by following established patterns. They began with someone refusing to accept that the established pattern was the right one.

This is not coincidence. When you map the cognitive profile of neurodivergence against the cognitive profile of innovation — non-linear thinking, pattern recognition in noise, hyperfocus on high-interest problems, resistance to convention, willingness to endure social friction in pursuit of an insight — the overlap is not incidental. The traits that the industrial economy treated as deficits were always, in the right context, assets. The context has now changed, dramatically and permanently.

The Honest Caveats Karp Does Not Always Make

I believe Karp is right. I also think he is telling a more convenient version of the story than the complete one, and honesty requires saying so.

Neurodivergent traits are rarely a menu from which you select only the attractive items. The same ADHD that produces divergent thinking and hyperfocus also produces executive dysfunction — genuine difficulty managing deadlines, organising multi-step tasks, and sustaining consistent output across work that does not trigger the reward system. The same autistic pattern recognition that makes certain people exceptional systems thinkers can come with real challenges in social communication and sensory processing. Dyslexic visual-spatial intelligence arrives alongside phonological processing difficulty that makes many standard professional environments exhausting to navigate.

The neurodivergent people who have succeeded in tech and innovation have done so, in the main, despite structural barriers — not because those barriers were removed. The employment rate for adults with ADHD in full-time work runs roughly 43 percent lower than for neurotypical adults. Estimates of unemployment among autistic adults run as high as 80 percent. The talent was always there. The environments were not.

Karp's Neurodivergent Fellowship is a meaningful signal, but a fellowship for elite builders at a defence-tech firm is not a structural answer to structural exclusion. If we are genuinely entering an era where non-linear cognitive traits are the highest-value human skill, the implication is not a fellowship. It is a wholesale rethinking of how we educate, credential, hire, manage, and support people whose brains work differently. We cannot celebrate the neurodivergent mind's output while continuing to make the environment in which it operates hostile.

AI as the Unexpected Equaliser

There is a second, underappreciated dimension to Karp's argument. AI does not only threaten the playbook-following mind. It also, quietly and powerfully, liberates the non-playbook mind.

The neurodivergent professional's historic disadvantage was not only that their best work was undervalued. It was that the administrative tax of operating in a neurotypical world consumed enormous energy that might otherwise have gone into the work itself. The effort required to produce an organised, formatted, sequentially coherent report; to manage a calendar with precision; to draft a perfectly conventional email; to navigate the unwritten social rules of a meeting — these are the tasks that drain a neurodivergent mind while costing a neurotypical one almost nothing.

These are also, almost without exception, the tasks that AI now handles effortlessly. The implication is significant. AI is not only eroding the value of neurotypical cognitive strengths. It is acting as a prosthetic for neurodivergent cognitive challenges — removing the friction that prevented the non-linear insight from reaching the world. The person who always had the vision but struggled with the plumbing of standard professional life now has a tool that manages the plumbing for them.

The Inversion Karp Is Actually Describing

Here is the reversal at the heart of this moment, stated plainly.

For a century, the economy spent enormous resources trying to make certain kinds of brains more like machines: more reliable, more sequential, more compliant, more standard. Pharmacology developed tools to reduce the divergence. Education systems developed behavioural frameworks to enforce conformity. Workplaces developed management structures to standardise output. The entire infrastructure of modern institutional life was, in a meaningful sense, a project to turn human cognition into something closer to what we now simply call AI.

We succeeded. We built the machine. And the moment we did, the value of the human qualities we had been suppressing inverted completely.

The person who could follow the playbook perfectly is now competing with a system that follows it infinitely better. The person who could never follow the playbook — who was too distractible, too non-linear, too obsessive about the wrong things, too insistent on seeing the problem differently — is suddenly not competing with anything. Because the machine, for all its power, cannot do what that mind does. It cannot decide the playbook is wrong. It cannot see the pattern that does not yet have a name. It cannot build the thing that has not yet been imagined.

"The irony of the century: the economy spent a hundred years trying to make people more like machines. The moment it succeeded, the machines arrived."

Karp is not predicting the future. He is describing the present. The neurodivergent mind was never broken. It was always running a different operating system. The modern economy just took this long to need it.

What Comes Next

I am not suggesting that every neurodivergent person is about to become a Palantir Fellow, or that the labour market disruption bearing down on millions of people is secretly good news in disguise. It is not. The displacement is real, the hardship will be real, and the people most exposed are not always those best positioned to adapt.

What I am suggesting is that the underlying logic of value creation is shifting in a direction that has been a long time coming. The Gartner research projecting that one in five Fortune 500 sales organisations will actively recruit neurodivergent talent by 2027 is not a data point about charity. It is a data point about competitive advantage. The Harvard Business Review analysis finding that teams with neurodivergent members showed measurable gains in productivity and error-detection is not a feel-good story. It is evidence about what kinds of cognition are actually useful when the routine work has been automated away.

The institutions that figure this out earliest — the companies, the schools, the governments — will have access to a pool of cognitive talent that their competitors are still filtering out at the front door. In an economy where the only durable human edge is the ability to see what machines cannot, that may turn out to be the only advantage that endures.

Alex Karp is an unusual person making an unusual argument in an unusual way. He is also, in this case, correct. The playbook is dead. The people who never learned to follow it are, at last, no longer at a disadvantage for that reason.

Some of us have been waiting a very long time for that sentence to be true.