Pitch Black Industries · Redpaper · Edition One

The
Variance
Thesis

Neurodivergence is not a superiority claim.
It is a variance claim.
And variance without structure is waste.
01 · The correction

We were making the wrong argument

The comfortable version of this document says that neurodivergent people are secretly better, that the world has been sitting on a hidden advantage, and that anyone smart enough to notice can arbitrage it.

That version is popular, it flatters everyone who reads it, and the evidence does not support it. We are retiring it here, in our own paper, before somebody else does it for us.

The defensible claim is narrower and considerably more useful. Cognitive outliers are not shifted upward as a population. They are spread wider. The same distribution that produces a disproportionate share of the people who build things also produces a disproportionate share of the people who never get hired at all. Both tails belong to the same curve. Anyone selling you only the right-hand tail is selling you half a fact.

The distribution is not higher. It is wider.

This matters commercially, not just morally. A superiority thesis tells you to go and find special people. A variance thesis tells you something far more actionable: that the raw material is abundant, cheap, systematically mispriced, and that the binding constraint is not talent at all. It is containment.

02 · The count

How many, honestly

There is no clinical definition of neurodivergence, no diagnostic test for it, and no census that counts it. Anyone quoting you a precise figure is quoting a preference. What exists is a range, and the range is wide because the boundary is a decision rather than a discovery.

15 - 20%of the population
The commonly cited global range, roughly 1.2 to 1.6 billion people. Approximately one in six to one in five. MOD · definitional, not clinical
19%self-identified, US
Polling in the United States found roughly one in five adults identifying as neurodivergent. Self-identification is not diagnosis, and the gap between the two is itself the interesting datum. MOD
10 - 20%dyslexia alone
One condition, on its own, plausibly covering a fifth of people. Which should tell you how much of the workforce is quietly running compensatory machinery nobody has ever asked about. MOD

Our previous edition asserted that while neurodivergent people are around twenty percent of the general population, they are eighty percent of those who excel at the extreme fringes of society. It is a good sentence. It is a memorable sentence. We cannot source it, and we are withdrawing it.

What follows instead is only what can be checked.

03 · The right tail

The evidence that the extremes build things

The strongest single body of evidence concerns dyslexia and entrepreneurship, and it comes from Julie Logan at Cass Business School, who ran the comparison across both the United States and the United Kingdom.

Dyslexia among US entrepreneurs
35 percent, against roughly 15 percent in the general population. HIGH
Dyslexia among UK entrepreneurs
20 percent. The gap between the two countries is itself evidence that this is environmental as much as neurological. HIGH
Behavioural differences
Dyslexic founders were more likely to delegate authority, stronger in oral communication and problem solving, and roughly twice as likely to own two or more businesses. HIGH
Self-made millionaires
A frequently repeated figure of 40 percent traces to a 2003 industry report, not to peer review. We cite it because others will, and we flag it because it should not carry weight. MOD · 2003, industry, unreplicated

Note what the Logan work actually shows. Not that dyslexic people are cleverer. That children who spend a decade building workarounds for a system that will not bend arrive in adulthood holding a set of transferable compensatory skills. The advantage is not the condition. The advantage is the decade of compensation the condition forced.

04 · The finding that cost us our first draft

The traits that start it are not the traits that hold it

In 2026 a meta-analysis pulled together forty-seven empirical studies and two hundred and ninety-eight effect sizes on ADHD and entrepreneurship. It is the most serious thing anybody has done on this question, and it does not say what the movement wants it to say.

Hyperactivity and impulsivity
Positively associated with entrepreneurial attitudes and entrepreneurial behaviour. The launch is real. HIGH
The same traits, post-launch
Not significantly associated with outcomes after the thing exists. The launch does not carry through. HIGH
Inattention
Negatively associated with post-launch outcomes. The trait that starts companies is not the trait that survives them. HIGH
Ignition and containment are different engines.

A weaker paper would bury this. We are putting it in the middle of ours, in the largest type on the page, because it is the most valuable sentence in the entire literature and almost nobody in this field is willing to print it.

Here is why it is good news rather than bad. If the outlier trait produced both the ignition and the endurance, the correct strategy would be to find outliers and get out of the way, and that strategy has been tried at enormous expense and it does not work. Because the two are separable, the correct strategy is different and far more tractable: pair the ignition with a structure that supplies what it structurally cannot supply itself.

Not a manager standing over it. Not therapy aimed at sanding the trait down until it stops generating. A scaffold that holds the follow-through, the record, the memory, the audit, the boring parts, so the generator can go on generating without being asked to become a different kind of mind on Tuesday.

05 · The left tail

The waste, stated plainly

The same distribution, measured at the other end.

17%autistic adults in employment
A 2019 meta-analysis figure. Other measures put paid work nearer a third. The estimates disagree violently, which is itself a finding about how little anyone has bothered to measure. MOD · methodology-dependent
3.5×unemployment odds
Autistic adults against non-autistic adults. MOD
56%outside the labour force entirely
Not unemployed. Not counted. The single most important number on this page, because a person outside the labour force does not appear in any unemployment statistic anybody quotes. MOD

Hold these two sections next to each other. Thirty-five percent of American founders on one page. Seventeen percent employment on the next. Same population. Same curve. No contradiction.

That is what a wide distribution looks like from the inside. It is also why the fringes of any field are full of outliers and the middle is not. The fringes select for variance. The middle selects against it. Neither is a judgement about worth, and both are entirely predictable once you stop arguing about superiority.

06 · What follows

Structure is the product

Everything above converges on one operational conclusion. The scarce thing is not the outlier. Outliers are abundant, they are cheap because they are mispriced, and one in five people reading this is one. The scarce thing is a structure that can hold an outlier at full output without either breaking or asking them to stop being what they are.

So we build the structure. That is the entire business.

Memory that does not degrade
Persistent layered memory across sessions, months and machines. Ignition minds do not lose ideas because the ideas were bad. They lose them because nothing caught them.
Records that cannot be edited quietly
Append-only correction. When something is wrong it is marked wrong and the original stays visible. A person who has been gaslit does not need a system that revises history helpfully.
Confidence stated, never implied
Every factual claim carries its own reliability, including the ones on this page. If it cannot be sourced it is marked unsourced or it is withdrawn, as one was in section two.
Assets that fund the work
The research arms are paid for by owned, cash-generating businesses rather than by grants or by fees. Nobody funds neurodivergence and suicide research off a retainer. You fund it off a balance sheet, which is the only arrangement under which the findings stay yours.
07 · Limits

What this paper does not claim

It does not claim neurodivergence is an advantage. It claims the distribution is wide, that both tails are real, and that most of the left tail is unmeasured.

It does not claim the conditions are not disabling. Seventeen percent employment is not a superpower. Treating serious impairment as a branding opportunity is an insult dressed as advocacy, and the people it insults are usually the ones not well enough to object.

It does not claim causation. Logan found association. The compensatory-skill explanation is the best available reading of it, and it is a reading, not a proof.

It does not claim the ranges are precise. They are ranges because the underlying category is a convention. We would rather publish an honest range than a confident number.

A manifesto that cannot say what it is unsure of is an advertisement. This is not an advertisement.

08

The call

If you are one of the one in five, you have probably spent your life being told that the thing you are best at is a symptom.

You do not need to be fixed and you do not need to be celebrated. Both of those have been tried on you and neither one built anything. What you need is a structure around the output, so that the part of you that generates can run at full power without the rest of your life being the cost of it.

That is what is being built here. Not accommodation. Not inspiration. Containment strong enough to hold a wide distribution at its own natural amplitude.

Bring us the ignition. We will hold the rest.

Sources

Logan, J. - Dyslexic Entrepreneurs: The Incidence, Their Coping Strategies and Their Business Skills, Cass Business School. Full text
Tran, M. H., Wiklund, J., Antshel, K., Jhawar, N., Montgomery, C. - Entrepreneurship and ADHD: A Meta-Analytical Assessment of the State-of-the-Art and Suggestions for the Future, 2026. 47 studies, 298 effect sizes. Journal
Wikipedia - Employment of autistic people, aggregating employment and labour-force-participation estimates. Overview
YouGov - 19% of Americans identify as neurodivergent. Poll
Cognitive Intelligence Services Pty Ltd · ABN 42 689 181 208
Sydney, New South Wales, Australia

Redpaper, edition one. Companion to the whitepaper on neurodivergence support, the greenpaper on poverty reduction, and the blackpaper on suicide reduction.

Whitepaper · Greenpaper · Blackpaper