Afrikaans
Akan
Albanian
Amharic
Arabic
Armenian
Azerbaijani
Basque
Belarusian
Bemba
Bengali
Bihari
Bosnian
Breton
Bulgarian
Cambodian
Catalan
Cebuano
Cherokee
Chichewa
Chinese (Simplified)
Chinese (Traditional)
Corsican
Croatian
Czech
Danish
Dutch
English
Esperanto
Estonian
Ewe
Faroese
Filipino
Finnish
French
Frisian
Ga
Galician
Georgian
German
Greek
Guarani
Gujarati
Haitian Creole
Hausa
Hawaiian
Hebrew
Hindi
Hmong
Hungarian
Icelandic
Igbo
Indonesian
Interlingua
Irish
Italian
Japanese
Javanese
Kannada
Kazakh
Kinyarwanda
Kirundi
Kongo
Korean
Krio (Sierra Leone)
Kurdish
Kurdish (Soranî)
Kyrgyz
Laothian
Latin
Latvian
Lingala
Lithuanian
Lozi
Luganda
Luo
Luxembourgish
Macedonian
Malagasy
Malay
Malayalam
Maltese
Maori
Marathi
Mauritian Creole
Moldavian
Mongolian
Myanmar (Burmese)
Montenegrin
Nepali
Nigerian Pidgin
Northern Sotho
Norwegian
Norwegian (Nynorsk)
Occitan
Oriya
Oromo
Pashto
Persian
Polish
Portuguese (Brazil)
Portuguese (Portugal)
Punjabi
Quechua
Romanian
Romansh
Runyakitara
Russian
Samoan
Scots Gaelic
Serbian
Serbo-Croatian
Sesotho
Setswana
Seychellois Creole
Shona
Sindhi
Sinhalese
Slovak
Slovenian
Somali
Spanish
Spanish (Latin American)
Sundanese
Swahili
Swedish
Tajik
Tamil
Tatar
Telugu
Thai
Tigrinya
Tonga
Tshiluba
Tumbuka
Turkmen
Twi
Uighur
Ukrainian
Urdu
Uzbek
Vietnamese
Welsh
Wolof
Xhosa
Yiddish
Yoruba
Zulu
Wouldn't it be nice to say, no, there's nothing wrong with you.
And in fact, you know,
that we need exactly your type of personality in order to be,
you know, creative and inventive as a community.
That sounds lovely,
doesn't it?
It's the sort of thing you'd want someone to say to you on a bad day,
to cheer you up a bit.
But let me tell you what she means by it.
She means that people like me,
late identified autistic adults,
like you,
The ones who can hold a conversation and get the joke and mask our way through a birthday party perhaps.
We should be taken out from the autism diagnosis she thinks.
Reclassified as having an interesting personality instead of a neurotype.
She believes the kindest thing she can do is to look a late diagnosed person in the eye
and tell them that there is nothing wrong with them and in the same breath
take away the one word, the one label that finally makes the whole life make sense.
This is not some angry bloke with a ring light and a grievance.
This is Dame Uta Frith, one of the most important autism researchers who has ever lived.
And in May she sat down with a new scientist and made the case that the autism spectrum has,
in her words, "come to its collapse".
She used another word in that interview,
she said it was dangerous to cling to the autism label.
Now hold that word because by the end of this video I'll be giving it back to her.
So today we are taking her argument apart,
all of it, every piece, and I'm going to be scrupulously
fair to her first because she has earned that,
and then I'm going to put her arguments in the
litter box,
I'm going to feed them to the tigers,
that sort of thing.
Anyway,
hey friends,
I'm Mike,
a late identified autistic adult.
Grab a cup of tea for this particular video,
I have got a huge
flask of tea,
acquire a cat or a plushie and settle in,
because this one is a big one and
it matters more than most.
New Scientist then.
If you don't know it,
it is one of the oldest and most widely read popular science magazines
in the world.
It's been going since the 1950s, it is not a tabloid and it is not a scientific journal.
It sits sort of somewhere in the middle.
It's the place a curious person might go to find out what scientists are saying.
When something runs in New Scientist, it carries a certain weight.
People tend to take it as fact,
as settled, or at least close to it.
And in May they ran a long,
friendly,
admiring interview with Dame Uta Frith.
We'll get to exactly who she is in a moment,
but the short version first is that she is
one of the most influential autism researchers who has ever lived.
Genuinely a giant in the field.
So this is not a fringe figure in a fringe outlet.
This is one of the biggest names in autism science,
in one of the biggest science magazines
in the world.
And here is what she used that platform to say:
that the autism spectrum,
the whole idea
of it,
has, in her words, come to its collapse.
The people getting diagnosed today,
especially women,
especially adults,
have, and again her
words, absolutely no overlap with what she considers to be real autism.
TM.
So why am I making this video?
Three reasons.
One,
I think that she's wrong and not wrong in some vague way that people might say on the internet,
oh, you're wrong,
you know, without specifying reasons,
but wrong in a way that I can show you piece by piece using her own words and the actual evidence.
Two, because of who she is,
where she said it,
this is not going to stay in a magazine interview.
It's going to move through teachers and clinicians and policymakers and doctors,
etc.
at the exact moment that this country is deciding how much support autistic people are allowed to have.
And three,
because I am one of the people that she's talking about.
I am late identified,
I can speak,
and the kind that she thinks doesn't really count.
I've got a horse in this race and I'm not going to pretend otherwise.
That's my bias for what it's worth.
If you are autistic,
you owe Uta Frith a debt,
and I want to say that plainly before I disagree with a single word.
When Frith started out back in the 1960s, the going theory for what caused autism was the refrigerator mother.
The idea,
and I do wish I was inventing this,
was that children were autistic because their mothers were cold.
They were emotionally frozen and unavailable,
you know,
unloving.
So you did not just have a child who was autistic and struggling.
You had a mother who had been told it was all her fault because they didn't love the child enough.
They didn't love hard enough. They were too cold.
they're frigid,
you know. Friff helped end that in a big way.
She sat with those mothers and she
looked at the evidence and she said no,
this is biological,
this is the brain.
That mattered and
it took a lot of nerve.
She went on to co-author the famous Sally Ann experiment,
you know,
the one that gave us the idea of theory of mind deficits in autism.
She brought Hans Asperger's
work to the English-speaking world.
She literally translated it.
And she taught, mentored, or shaped
a huge slice of the autism researchers currently working today.
Simon Baron-Cohen, the chap whose
name is on half of pretty much all autism research that you've ever heard of, was her PhD student.
So when I tell you that I think Uta Frith is wrong,
understand what I'm doing here. Just sort of like
swatting away some sort of crank with a podcast.
I am disagreeing with one of the architects of the
entire field.
That is a heavy thing to do and I do not do it lightly,
which is
probably why this video is going to be very very long.
Before I show you where
she goes wrong I'm gonna show you where she's right,
properly right, because if I
can't do that I've got no business making this video.
Four things on all four she
has a point.
Number one:
that autism is a wildly,
almost absurdly varied.
A non
speaking autistic person who needs round-the-clock care and an autistic
adult holding down a job while quietly white-knuckling their way through every
meeting are both on paper the same diagnosis,
same label.
Now that's an
enormous range to hang on one word.
Researchers like Lombardo and Mandeli
have written exactly this:
that autism is heterogeneous at every level from the
genome all the way up to behavior.
Frith is not imagining it,
she's not lying
when she says it,
it is real.
Number two:
there is a very real fear that people with the very
highest support needs get forgotten.
The non-speaking,
the ones who need full-time care.
In public
conversation increasingly shaped by those of us who are verbal and online,
and let's be honest,
never knowingly under-opinionated, there is a worry that their needs might slide down the agenda.
Now, I have my problems with how that worry gets packaged online,
and I'll get to it later,
properly because it matters.
If you want to know now the TLDR,
perhaps watch my levels video
linked above my head.
But the fear underneath it is fair.
High support needs autistic people
deserved not to be an afterthought.
What I will not accept and what I will not spend this video
dismantling is the idea that the way to protect them is to throw the rest of us out of the category.
Those two things have nothing at all to do with each other and she keeps acting as though they are
the same move.
We'll get to it.
At the start of this video there is an awful lot of we'll get to it,
but I promise you we will.
Number three:
she is right that there is no biological test for autism.
There is no blood test,
no brain scan,
no single gene that you can point out and say "there,
that's the one,
that's the autism gene,
that's the autism region of the brain"
or whatever.
We diagnose it by watching behaviour and listening to a life.
And yes,
that is scientifically awkward.
It would be tidier if they were a marker and you could have a blood test.
There is not.
Four.
That resources are finite.
Waiting lists are real.
The money for support is not infinite and working out who gets what is hard,
it's a material and it's a very miserable problem.
She's not wrong that the system is straining.
This is the thin strip of solid ground.
That
is the only thing that she is standing on.
In May 2026,
New Scientist ran a long interview
with Frith, written by the journalist Thomas Lewton and filmed by David Stock.
A sit-down
in her living room.
Tea.
Homemade macaroons.
Model trains on the shelves.
Very quiet and
uncivilised.
And it was not the first time because back in March she had said much the same thing to the TES
or the Times Educational Supplement.
That's a magazine that goes out to school teachers under the headline
"Why I no longer think autism is a spectrum". Hold that detail,
teachers,
we're going to come back to it.
If you're taking notes, and I know some of you will be,
we'll close all of those loops towards the end.
Here is the heart of it in her own words.
It just becomes a question of why one should cling to the label autism.
And I would say it's dangerous to do that because it completely ignores the profound autism,
which has these very severe communication difficulties.
So I'm still not understanding what the essential problem with calling both autism is.
Is it just that you see absolutely no overlap?
Absolutely no overlap,
yes.
And I'm really,
in a sense,
completely astonished that people want to cling to this label
when it, you know, has outlived its purpose.
And you really should look for new labels and new
categories,
new ideas.
Collapsed.
No overlap.
Outlived its purpose.
Nobody reaches for language like
this when they want to tidy a category up. They reach for it when they want it gone.
Now you might be thinking, so what? An academic has an opinion,
academics disagree for a living,
why make a 40 minute video about one interview?
Because of who she is and because of where
this lands.
If I were to say the spectrum has collapsed,
nothing happens.
I am a bloke
on YouTube with 16 cats and a strong opinion.
But when Uta Frith says it, it does not stay
in a magazine.
It travels,
it gets cited,
it ends up in policy submissions,
in government
reviews in the mouths of people deciding how much support an autistic child is allowed
to have.
Her name carries weight and weight moves things.
And remember where she said it first:
the TES.
Two teachers.
At the exact moment the UK is fighting over special educational needs support and whether
to cut it back.
Right now in England,
around a quarter of a million people are sitting on a waiting list
just to be assessed for autism.
Not treated,
assessed.
and roughly 9 in 10 of them have been waiting longer than the period the NICE guidelines
say that anyone should.
The National Autistic Society puts the average wait at around 17 months.
At the same time,
the number of education,
health and care plans citing autism has nearly
tripled in a decade
from around 54,000 to 149,000,
according to the Institute for Fiscal
Studies.
one rise accounts for 40% of the entire increase in EHCPs across the board.
I know what some of you are thinking because it is the obvious objection.
That is,
the same rising number that Frith reaches for.
She looks at that climb and sees a bubble,
a fashion,
people being swept in who do not belong.
So why do I get to look at the identical graph and see something completely different?
Fair question,
and here is the answer.
When you widen the criteria and train more clinicians and finally start looking for autism in
women particularly and adults instead of just little boys,
the number going up is exactly what
you would predict.
This is not an epidemic,
it is not a surprise,
it's a backlog clearing if anything.
The boring explanation is that detection catching up with reality fits the data at least as well as
hers and does not require you to believe that hundreds of thousands of people and their clinicians
collectively kidding themselves.
A quarter of a million people queuing for a door that barely
opens,
a support system buckling under pressure,
a government hunting for reasons to tighten the
criteria.
And into that room walks one of the most respected autism researchers alive to announce
that the whole category might just be meaningless.
Now do you see why she is entitled to her opinion
and doesn't quite cover it?
There are two more things that she said that I keep circling back to.
We have to really rethink the whole idea.
And as far as I can see, nobody is really brave enough to do that.
The first is that she casts herself as the brave one,
the lone voice willing to say the unsayable.
She says that she's the only one brave enough to tell the truth.
She says that the label is the dangerous thing.
So let's test both of those properly against the
actual science.
And to do that,
we have to start with the single idea her entire argument is built
on
one idea from 1934 that her own profession quietly threw in the bin about 50 years ago.
Here is the load-bearing wall of Frith's whole case.
Pull this out and the rest comes down, so watch closely.
She almost says it in passing.
The interviewer puts it to her that autistic women often mask,
and that they camouflage their autism to get through the day,
and this is why so many were missed.
And here is her answer:
And as researchers like Gina Rippon have said,
that there are coping strategies that autistic women have because of the pressure to maybe
socialize more than men do and so that you end up with you know you end up masking or camouflaging
so you you may present as having quite normal social behaviors but actually there can be
different kinds of neural processes going on which can also lead to internal stresses and
experiences?
What's your feeling? Well, I think it creates an extraordinary paradox,
so that you can diagnose autism if you show the symptoms,
and you can diagnose autism
not showing any of the symptoms,
but assuming that there is this masking going on.
So how can you ever falsify a diagnosis?
It just doesn't make sense to me.
And if you can never prove it wrong,
it's not science.
Therefore, she says masking is not real.
And the women who rely on it to explain their late diagnosis are not really autistic.
Now, I want to be fair because that idea has a long and respectable pedigree.
It has a name.
It is called falsifiability,
and it comes from a philosopher called Karl Popper,
who set out in 1934.
Popper was wrestling with a real question.
How do you tell science from fortune telling?
and his answer was elegant.
A scientific claim is one that sticks its neck out.
It forbids something.
Saying "the sun will rise tomorrow"
is science because if the sun does not rise,
you are wrong.
Mercury being in retrograde and making you tired is not because whatever happens,
the horoscope was right all along.
It's a lovely idea and it caught on for a while and I understand
completely why she reaches for it,
but there is one problem.
Philosophers of science spent the
next 50 years taking Popper's ideas apart,
and by now naive falsifiability is not the test anyone
serious uses,
and I do mean anyone.
Here's the first crack, and it has a wonderful name.
It's the
Duum-Quine thesis.
It says that you can never actually test one idea on its own.
Every experiment
drags along a whole rucksack of assumptions along with it.
So when a prediction fails,
the failure never tells you which part broke.
Maybe the theory is wrong,
maybe your instrument was miscalibrated, or your sample was off,
or one of the 10 background assumptions that you didn't even notice,
maybe one of those was the
faulty one.
When astronomers found a planet in the wrong place,
they didn't just bin the theory of
gravity,
they went looking for an unseen planet that must be pulling on it.
And that's how they
found Neptune.
The anomaly was real.
Throwing out the whole theory would have taken the wrong move.
And here is the crack that finishes the job. A philosopher called Larry Laudan showed that
Popper's test fails in both directions at once.
Astrology is perfectly falsifiable.
You can check
its predictions and watch them fail.
Meanwhile,
large chunks of respected theoretical physics
are not clearly falsifiable at all,
but nobody is throwing physicists out of the building.
The falsifiability ruler can cut the wrong things.
It can let nonsense in and it can throw good
science out.
Uta Frith is judging masking with a ruler that her own discipline measured,
found wanting and is quietly retired half a century ago.
Psychology is often called a soft
science and excluded from other STEM disciplines for this reason.
It is a bit fluffy.
The standard that psychology actually runs on is called construct validity and it was set out by
two researchers in 1955.
The idea is simple and rather beautiful.
You cannot see intelligence.
You cannot see depression or anxiety or working memory.
You cannot hold any of them in your hand.
So how do you know that they are real?
You check whether they behave the way a real thing should.
Do they show up consistently?
Do they predict other things?
Do they hang together in a web of of relationships that makes sense.
If a thing converges with what it should converge with,
separates from what it should separate from,
and predicts what it should predict,
then it is a real measurable construct.
Not because you falsified it in one dramatic experiment,
but because it earns its place in the network.
And by that standard, the one that the entire field
uses,
masking is not some vague feeling.
It is one of the best behaved constructs in autism research.
We have a validated measurement tool for it.
It's called the Camouflaging Autistic Traits
Questionnaire or the CATQ.
I made a video about this right here if you want to watch it. It was
built by Laura Hull and colleagues and it converges with other autistic trait measures exactly as it
should.
It tells autistic and non-autistic people apart exactly as it should and it predicts things,
serious things.
Higher masking predicts higher anxiety,
obviously.
Higher masking predicts
depression.
Autistic burnout,
which Jura,
Raymaker and colleagues mapped out in detail,
and in the work of Sarah Cassidy and her team, camouflaging predicts suicidality.
Before anyone says, ah,
but that's just people filling in a questionnaire online about their
own feelings.
No,
Meng Chuan Lai and colleagues put autistic women in a scanner and found
distinct neural patterns associated with camouflaging.
So this is not just self-report.
There is something
measurably different happening in the brain when an autistic person masks.
That can leave
a fingerprint.
Note that is different from detecting autism on brain scans as detecting
an autistic behaviour on brain scans.
Different thing.
Okay, just putting that out there as a little footnote.
Let us be very clear about what Frith has done.
She's called masking unfalsifiable and therefore unscientific,
while sitting on top of a lot of literature that operationalises it, measures it, scans it,
and uses it to predict whether an autistic person will still be alive in the future.
The thing that predicts whether or not you'll still be alive is the precise opposite of untestable.
But here's the part that turns this from a disagreement into something that's a great deal more uncomfortable.
The moment you pick up Frith's ruler and apply it honestly,
evenly,
to everyone,
the first thing that it condemns is not masking.
It is her.
Her single most famous contribution is Theory of Mind.
Sally has a basket and has a box,
and Sally has a marble and puts it into her basket and covers it up.
the idea that autistic people struggle to track what other people are thinking.
The Sally Ann test.
It's in every textbook.
It is,
more than anything else,
the thing that her name is attached to.
Listen to how she describes it now in the same interview.
We say the particular difficulties in social communication,
which are only one aspect of autism,
are possibly rooted in a difficulty in mentalising.
That was an idea that we could have turned out completely wrong.
Could have turned out completely wrong.
Only true for a small subgroup.
And she is not wrong to hedge,
because independent researchers have taken theory of mind apart
with exactly the kind of rigour that she demands of masking.
Morten-Anne Gernsbacher and Melanie Youngu went through the literature in 2019 and found three
things:
the original findings often fail to replicate,
the tasks correlate with how good
your language is,
not whether you are autistic,
and plenty of non-autistic people fail while
plenty of autistic people pass.
They also found that something like three-quarters of the most cited papers on the subject just
assert the deficit as fact without offering any new data to back it up.
They just go,
"Uter Frith says autistic people lack theory of mind, I rest my case", right?
It's just an idea,
repeated,
until it sounded true.
And then there's the problem that should really give Uta pause,
and it comes straight out
of her own framework:
psychopaths.
People with psychopathy tend to do perfectly well on the cognitive side of theory of mind,
the part that these tasks actually measure,
working out what someone else knows or believes.
Where they differ is affective empathy,
the part where you feel anything about it.
And that split is the whole problem for this test.
If you can pass a mentalising task while feeling nothing for the person in front of you,
and yet a deeply empathic autistic person can fail it, then the task is not measuring care
or connection or whether someone is a good person to have in your life,
it's measuring
one very narrow cognitive operation,
and Frith built an entire theory of autism's social
heart on top of it.
There is a better way to think about social difficulty,
and it has a name.
The double empathy problem set out by Damien Milton in 2012.
The idea that the breakdown in understanding between autistic and non-autistic people is
not a fault inside the autistic person,
it runs both ways.
Non-autistic people are just as bad at reading autistic people as the other way around.
From an empathy perspective,
if you put autistic people together,
they communicate just as
well as if you put neurotypical people together and non-autistic people together.
The mismatch lives in the gap between two different ways of being,
not inside one broken
brain.
And here's the really lovely bit.
You don't have to take my word that any of this is mainstream now because the new scientist
article you're watching clips from says it too.
It notes in black and white that the lack of empathy idea is disputed and that autistic
people communicate effectively with each other.
Her own venue concedes the point.
And then right at the end of the interview,
the journalist asks her a direct question.
Of all the theories of autism that have been put forward over the decades,
which ones does
she think has stood the test of time?
There have been many theories of autism,
of what autism might be put forward over the
decades.
Which do you think have stood the test of time?
Very little has stood the test of time because I think we do know too little about brain
and mind and that is still an area that is in a sort of early stage and a lot of things
are possible with new techniques.
Now remember who's answering.
This is the woman whose theories are the theories.
Theory of mind, weak central coherence.
If anyone's work is in that pantheon,
here it is.
She's not only talking about her own work, she's talking about the whole field.
The whole field.
Autism.
The whole autism,
understanding of autism, autism research a lot.
For this stuff,
it is substantially is, actually, her work.
You cannot say very little in autism theory has survived and then carve out an exception
for the two biggest theories in autism,
both of which have your own name on them.
She is including herself.
She has to be.
And that's not me being unkind to an 80-something-year-old woman.
Those are her words in her chair over her macaroons in an answer to a direct question about what survived.
Uta Frith demands an objective,
falsifiable,
biomarker-backed standard from masking and
from the women who describe it.
Her own life's work, by her own admission,
in the same conversation,
never reached that
standard.
theory of mind work might be,
is mostly wrong.
The imaging,
very little of that has survived.
She demands of us what she admits she never delivered.
Twenty odd minutes into a video about the philosophy of science and a woman's cardigan,
you are exactly my kind of person.
So do me a favour,
tap the like button.
I know, I know,
but this is the honest truth of how this works.
YouTube decides whether to show this video to the next late diagnosed person who needs
it almost entirely based on whether you
yes you watching this video
are engaging with
it.
So a like, a comment, a share to one person who would care.
It costs you nothing and is the difference between this reaching 10 people and 10,000
people.
She is pointing her arguments at women,
specifically autistic women diagnosed late.
Here's her case.
She says that there are two groups:
the early diagnosed group,
which she describes as severely
affected with intellectual disability and language delay,
and the late diagnosed group
who she describes like this:
Suddenly increasing number of cases in these late diagnosed, especially women,
is they have
this wonderful communication.
They have absolutely no difficulty in language.
They have everything.
They are,
you know, they still want to be autistic or they are labelled autistic,
but I wonder whether they have, they would be much better recognised and much better treated and understood if one looked at what really their problems actually were.
The problems are not the same as of those other cases.
And as you say, you could call it type 2 autism or whatever it is.
And then she reaches for a number.
She says that in the late diagnosed group,
the old four to one ratio of boys to girls collapses
and you get almost as many women as men.
And to her,
this is a smoking gun,
she says.
The classic proportion of male to female
in autism was four to one
and quite soon changed to three to one.
So preponderance of males,
to reach three boys to one girl.
And that has remained amazingly stable in the early diagnosed group.
But in the late diagnosed group, we see a completely different pattern.
We suddenly see more and more females appearing.
And this has been shown in another study,
very recently published this year,
a study of the Swedish population,
which showed that after the age of 15,
when you get diagnosed
at that point,
you get practically as many girls as boys.
In other words, it's a huge increase in girls.
But she is using data from a very, very large Swedish population study published in the BMJ this year.
And it does not say what she needs it to say.
It actually says the opposite.
The researchers and the editorial that ran alongside it concluded that the narrowing ratio
from 4:1 to 3:1 to... and so on is evidence that autistic women were systematically missed
for decades and are finally being found.
The editorial says it in plain English. It describes
late diagnosed women as, and I quote, "just as autistic as their male counterparts".
Oh well,
the CDC's own American data shows the same ratio narrowing over time.
The most boring,
most parsimonious reading of that number is not "We let the women in!
We're letting the wrong women in!"
No,
it's actually "We kept the right women out and we're only now letting them in through the door."
So she's taken a finding about women being under-recognised and read it as women being
over-included.
Same number,
opposite conclusion.
My conclusion is the same as what the study says
And I wonder if she's actually reading properly.
It just makes me wonder if Uta Frith is such a powerful mind and is she even reading?
And here is the mechanism that she won't look at, which is the one that explains the whole thing.
Masking,
which remember we have just established is real,
measured and visible in a brain scan.
Autistic girls grow up under far more social pressure to fit in than boys typically do.
so they tend to learn earlier and harder to camouflage.
Not that they're hard to camouflage,
just they learn to camouflage harder with more effort,
if that makes sense.
We're not camouflaging
girls right now.
They have it hard enough as it is.
They learn to script the conversations,
right?
Copy the popular girl, that sort of thing. Hold it all together all day at school and fall apart
the moment they get home.
And because the autism textbook was written around boys,
a girl who masks
well looks,
well to a clinician,
like a girl who is just fine.
The masking is not a footnote,
it's the reason that the four to one ratio exists in the first place.
And then Frith goes further,
and this is the tool that made me put my tea down.
I think it is certainly is, you know,
a research area to really go into and answer these questions,
but you have to face the fact
that these late diagnosed people had a perfectly unremarkable childhood.
They developed quite normally,
at least, you know, from what one basically hears.
I believe also that it's very easy to elicit memories or reports of slightly unusual behavior
if you want to do that.
and you can get basically anybody,
including you and me,
to have some episode in your early life
that might say, that looks a little bit autistic.
She's saying that late diagnosed people had,
in her worlds,
a perfectly unremarkable childhood and that autistic memories they describe are
basically suggestible noise.
That if you go looking,
you can talk anybody into believing
that they are a bit autistic as a child.
But think about what that actually is.
That's not data,
that's just her opinion,
it's just an assumption, it just wears the costume of
data.
You know, an Afghan flavoured lab coat,
if you will.
You assume that the
childhood was typical and then you treat every piece of evidence to the contrary
as the person fooling themselves.
It's a closed loop,
no one can win, it's circular
reasoning,
nothing can ever get it.
This is the ground that researchers like
Gina Rippon have spent their careers,
this is Gina Rippon there, have spent their
careers reclaiming.
The lost girls of autism.
The ones in the field,
in Rippon's telling,
simply did not bother looking for, because the entire picture of autism was built around small
boys.
The work of Bargiella and colleagues documented exactly how these women slipped through,
masking and adapting,
and being told again and again that they seem fine.
The cost of all that
masking is not nothing,
whatever Frith says.
The research from people like Leigh and McQuaid
links sustained camouflaging to anxiety,
to depression,
to a damaged sense of who you even are,
and as I said before,
to suicidality.
That is the opposite of an interesting personality quirk.
This is people quietly drowning while performing fine.
Look at the trap that Frith has built,
maybe without meaning to.
She says that she will not trust
an autistic woman's account of her own life.
I think the subjective experience is,
It's very,
very tricky.
You do need to verify it.
You do need objective criteria.
This objective experience is very malleable
by cultural opinion,
by the way you understand yourself,
you have memories.
That is just not a solid enough foundation.
Frith wants objective,
external,
third-party verification
before she will even believe that you're autistic.
But the objective,
external,
third-party system
appealing to is the exact same system that was built around boys,
white boys,
and well-off families,
pretty much.
The exact same system that missed these women in the first place.
She's proposing that we re-examine the lost girls of autism using the precise instrument that lost them.
And look,
I have a stake in this and I'm not going to pretend otherwise.
I am one of the people she's talking about because I was identified as autistic as an adult.
Not because it was fashionable,
not because someone talked me into a memory,
but because I spent decades knowing something did not fit and finally found an explanation that fit it perfectly.
And I'm not an interesting personality variant, at least I don't think so.
Let me know in the comments.
But I'm autistic.
The system just missed me.
As it, of course it did.
I'm 44 years old.
Of course the system missed me.
And I'm not willing to let the same person who helped build that system stand in her living room,
or sit in her living room I guess, and tell me that I imagined the whole thing.
Frith does not just rest on a feeling.
She points to the science,
she says modern genetics is
starting to find real subtypes, distinct kinds,
and this supports breaking the spectrum apart.
We don't have an objective way of diagnosing autism.
Even though we believe there is a brain
base, you believe there's a genetic basis.
There's a lot of evidence,
yes, that this is definitely
the case.
And there has been this general trend in recent years of these subtyping studies where
you find,
or researchers find,
patterns in genes,
in brain activity,
and also kind of experiences
and behaviours.
And then the question is, are there multiple kinds of autism?
I have read those
studies many times. I have really taken them to heart. In fact I made an entire video taking them
apart and I'll put it on screen.
Watch it after this one.
So I'm not going in cold here.
The studies
are interesting and they're not fraudulent,
they're not junk,
but they do not say what she needs them
to say and the reasons are baked into how they were built.
Starting with the big one, the 2025 study
in Nature Genetics. It took over 5,000 autistic children and then used machine learning or AI
to sort them into four subtypes based on more than 230 traits each.
Those traits were mostly parent-reported,
which in practice means a parent sitting at a kitchen table
scorecarding their own child on a long questionnaire.
That is not a knock on the parents,
they're doing their best,
but it does mean that the raw material of the entire study
is merely a parent's perception of their child's behaviour,
and that is just filtered through whatever that person thinks that autism looks like.
Frith spends the whole interview telling us subjective report is too unreliable to trust
while leaning on a study built entirely out of it.
So one of those four subtypes is the interesting one for our purposes,
it's the largest group,
around 37%, and the researchers called it social and behavioural
challenges.
They wanted to write core autistic traits,
hit their developmental milestones roughly
on time,
good language but a stack of co-occurring conditions like anxiety, ADHD and depression.
If you're listening carefully,
that group sounds an awful lot like Frith's so-called second group:
the late diagnosed,
verbal, anxious ones who she wants to remove from the autism category.
And there are more problems.
The first is that every one of these studies is built on children.
The 2025 one runs from age 4 to 18,
the 2023 one before it in Nature Neuroscience was children and
young people too.
So whatever they find,
they structurally cannot speak to autistic adults.
Yet it is autistic adults that Frith's entire argument is about.
She's using research on children to tell grown women that they are not autistic.
The second is that these are not the neat,
separable boxes that the headlines promised.
As I said in that video,
these are fuzzy clusters,
not species of Pokémon.
People sit on the borders,
people shift between them as they grow and their lives change,
and the researchers themselves are the first to say so.
The lead author said in plain words that finding four classes in the data does not mean that there
are only four kinds of autism.
The team behind the age of diagnosis genetics described their own work
as showing the diversity found within autism,
not the exit from it. Who paid for a study does not
tell you whether it's true.
A good study funded by the devil himself is still a good study.
So everything I just said, the children, the parent reports, the clusters that cut across age and sex,
That stands entirely on its own,
whoever signed the checks.
The next part is not me proving the science wrong.
This is me explaining why the framing around it keeps coming from the same shape.
This is really worth knowing and it's worth talking about.
Both studies came out of the same machine.
They were funded by Safari,
the research arm of Simon's
Foundation,
and they run on Spark,
the Simon's Foundation's own enormous collection of autism data.
This is a heavily genetics-focused,
heavily resourced research world, and it has very few
autistic people anywhere near the steering wheel.
One of the figures circling it, Alison Singer,
sits on the SPARC advisory board and runs a foundation that takes Simon's money and is one
of the people who coined the term "profound autism". She has argued on record that you need to be able
to call autism a severe disorder with symptoms and comorbidities.
So this is not neutral scientific
ground.
This is a research world built largely around the idea of autism as something to be
screened for and ideally prevented.
That does not make the data wrong,
it does explain why the people
holding it keep reaching for the same conclusions and why Frith picking up on the framing arrives
where she arrives.
It's a cultural thing.
And I would say it's dangerous to do that because
it completely ignores the profound autism which has these very severe communication difficulties.
Ah, profound autism again.
Frith reaches for that term to make her case.
The idea that there is a group so severely affected that lumping them in with the late
diagnosed is an insult to them.
And I have a whole video on profound autism right here,
so let me tell you what that term actually means, because it does not mean
what she is using it to mean.
Profound autism was proposed by a report in the Lancet Commission back in 2021,
and it was never proposed as a diagnosis.
It is purely an administrative term.
You will not find it in the DSM, you will not find it in the ICD.
No clinician will ever write it on a form. It was meant as a flag for research
and for directing care to the people who need the most of it.
And here is how the Lancet defined it:
someone who needs round-the-clock supervision,
who cannot be left alone,
who cannot manage their daily basic needs.
And this line,
which I read out in my earlier video,
and I'm going to read it again here because it is the whole thing,
the Lancet says profound autism is,
and I quote, "defined not by autistic features,
but by intellectual or language disability."
I will say that again for those at the back,
not by autistic features.
Profound autism is defined
not by autistic features,
but by intellectual or language disability.
Profound autism,
the thing Frith is holding up as real,
the severe,
the proper autism that the late diagnosed are supposedly diluting,
is a category defined by how much intellectual disability and language disability sits alongside
autism,
not by autism.
And that is the trick at the very centre of her split,
whether she intends it
or not.
When she looks back at her early diagnosed group and sees something profoundly different from
her late diagnosed group,
a great deal of what she's actually seeing is intellectual disability
and language disability in the one group and not in the other.
Those disabilities are real,
they matter enormously,
support for them is important,
they can completely change what a
person needs from the world.
But they are not autism,
they are conditions that often travel
alongside an autism diagnosis,
strip them out,
look at the autism on its own,
and the
unbridgeable chasm that she keeps describing narrows to just a small crack.
To show you how badly that term fits the job she's giving it,
there are autistic people
who speak fluently,
who have extremely high IQs, who still need round-the-clock care because
of severe anxiety or epilepsy or any number of co-occurring conditions.
Under the Lancet definition,
those people are not profound.
And yet their autism,
by any ordinary meaning of the word, really is profound.
The label sails straight past them.
So it does not even do the one job that it was built to do.
It's not a measurement of how autistic you are, it's a measure of how much intellectual
disability you have.
And Frith is leaning on it as though those two things are the same thing.
They are not.
Do you notice the move Frith has pulled earlier in the very same interview?
She said she told us that she does not trust big data.
She said:
I'm a little bit sceptical of big data,
basically because it is very,
very noisy.
And you don't,
you also, the bigger the data,
the more likely it is that there are errors
in collection,
in the kind of tests that you use.
When you think of behavioural tests or cognitive tests, they will be minimal because,
you know, you can't do two days or three days worth of assessment for one of,
you know, one million individuals.
Very,
very noisy,
right up until it appears to agree with her,
at which point it becomes
the bedrock of her argument.
You do not get to disown the method and keep the conclusion.
The subtyping studies that she's leaning on are big data bottom-up studies,
the exact
kind that she told us, she doesn't believe.
She trusts the data when it flatters her and distrusts
it when it does not. That is not scepticism,
that is just shopping.
At the end of the interview Frith
says something that sounds almost humble.
We are all different in these respects.
So this would be
an identity and absolutely wrong to pathologise.
Now in these cases I would really wonder whether
we could take out that now called subgroup from the spectrum and say,
actually,
you just
have a somewhat eccentric,
wonderful personality.
There is no way that there is anything wrong
with you.
This is a central proposal,
that late diagnosed group does not have autism,
but it has instead,
in her words, an extremely useful personality variant.
I think a wrong turn has been the pathologizing of what I think are probably extremely useful
personality variants that I would call autistic-like.
Yeah,
yeah capitalism I guess.
I mean she would call it autistic-like. I mean honestly if I'd
known I'd have called the channel Autistic Like or Autism Light AF or something and just saved
everyone all the bother.
Anyway,
take them out of the category,
tell them kindly that there is nothing
wrong with them,
right?
That's what she wants to do. If you reclassify somebody who has support needs,
that has a disability,
you reclassify them out of autism,
you reclassify them out of support.
The diagnosis is not a sticker,
it is the key potentially that can get you workplace adjustments,
educational support,
clinical care,
the basic recognition that you're not broken.
There is nothing wrong with you,
it sounds like a gift right up until the moment it it becomes
so you do not qualify right.
That's the same sentence,
there is nothing wrong with you so you
do you not qualify?" She's not saying the critical part,
she's not saying the conclusion,
she's just
sort of sitting on the nice part with the macaroons.
But she does gesture at the real machinery
underneath even if she doesn't dwell on it. Can we not find a way to meet all of these people's needs?
Well we have limited resources,
you have to remember that.
They have to be,
you know, distributed
wisely and fairly.
And it is very,
very difficult to judge that,
I have to say.
That is not a neuroscience finding,
by the way.
Who deserves support is a political question.
It's politics.
It's a moral question, an ethical question,
but it is not a scientific question.
She keeps sort of smuggling it into the conversation
as though it were a fact about the brain.
It is not.
Whether you are autistic is one question.
Whether the state will pay for your support
is a completely different question.
And she has quietly fused the two together
so that the answer to the second can be solved
by changing the answer to the first.
And while we're here,
let us actually look at this terrifying drain on resources,
the late identified adults,
right?
Because I am one of them.
And I'll tell you what I got from my diagnosis.
I got a letter,
I got a report.
I got six documents actually,
a nice little package.
I got a conversation and I got a reading list.
I got basically permission to stop hating myself
for the things that I've been failing at my whole life
without knowing why.
And that is the typical story,
by the way.
The vast majority of late identified autistic people
are not asking for a care package
or a residential placement or a penny of anything.
They're asking to be believed and to be supported,
to get the odd reasonable adjustment at work,
the kind that costs an employer roughly nothing.
to stop being told that they are imagining it.
The image Frith is leaning on
of hordes of fake autistic adults
swarming a finite pot of money
and starving the truly needly
is almost entirely imaginary.
The people that she wants to remove from the category
are for the most part not the ones drawing on the resources
that she is so worried about.
So the sort of austerity-flavoured rationing argument
does not even do what she needs it to.
She's picking the wrong people's pockets
to solve a problem that they were never causing.
We should ask, by the way, where this framing comes from,
because Frith is not having this thought in a vacuum.
I've spent a lot of time on this channel
mapping the autism research machine,
the funding bodies,
the genetic first priorities,
the organizations with very loud opinions
and the very few autistic people anywhere near the top,
anywhere near the steering wheel,
right?
The same world that wants to subtype autism
for a, you know,
quote, "precision diagnosis and care"
is the world whose framing she has picked up and run with.
She says it's not pathology.
The people that she has just handed this argument to
think it is nothing but.
And the biggest tell for me,
the thing that should kind of make everyone pause
is that her own student broke with her on this.
Simon Baron Cohen on the BBC
agreed that the word autism
may have stretched a little too wide,
which is a reasonable thing to say,
and then said that there is no evidence of over-diagnosis.
The man who helped her build the framework
looked at the same situation
and could not follow her to the same conclusion.
When your most famous student plants a flag
short of where you are standing,
that is worth noticing.
Every objective bar that she set for the rest of us,
she's missed herself.
And she told you so on camera.
She demands of us what she admits she never delivers.
It's not a category falling apart.
It's a category finally being asked to account for reality
and straining because it was drawn far too small
in the first place.
And as you say, you could call it type two autism
or whatever it is,
but why not call it, you know,
find a name that actually is suitable
and recognises that this is a different,
a different thing.
Go right back to the beginning.
The very first person ever formally diagnosed with autism,
Donald Triplett.
The man Leo Kanner literally labeled case one in 1943.
What was he like?
He had a degree in French.
He had perfect pitch too.
He could multiply three digit numbers in his head in seconds.
He worked at the bank, the same bank for 65 years.
He loved traveling the world.
He was verbal,
gifted, independent,
and he lived a long and contented life on his own terms.
That's patient one.
That is the foundation stone of the entire diagnosis.
So the idea that real autism means a non-speaking child
with an intellectual disability
and that a verbal capable adult cannot be autistic
is just wrong.
It was wrong in the very first case in the file
and we did not drift away
from some pure original meaning of autism.
We just spent 80 years forgetting it
because intellectual disability
kept getting tangled up with autism
and confusing the whole picture.
We can separate the two questions
that she spent the whole interview blurring together.
Is autism a scientifically valid construct?
The answer is,
to the best of our knowledge,
yes.
By the same standard psychology uses
for every other construct it has.
And the standard her own theories struggled to meet.
Question two is,
who should the state pay to support?
And that is a political fight, a real and hard one,
but it's not answered by a brain scan
and it's not answered by deleting people from a diagnosis.
Why Yuta Frith is getting so involved
in this funding question and the politics of disability
at such a late stage in her life,
I do not know.
And in the end, she tells you all of this herself.
Right at the close, the journalist asks
whether we are at a turning point in understanding autism
and she does something I actually admire.
She refuses to oversell it.
I'm not really,
I don't want to pretend that I actually know anything
or have the kind of facts that I could be sure about.
I think it's all in flux,
sadly,
but that's science.
It's all in flux, I cannot be certain of anything.
And also,
this category should be broken up.
These people removed, the label retired.
You cannot build a demolition order on a foundation
you've told us is still moving.
Do you think this is a critical turning point for our understanding?
Or we're nearing a turning point in our understanding of autism once again?
I think we're at a low point,
not at a turning point.
I think it could be a time where we have to really rethink the whole idea.
And as far as I can see, nobody is really brave enough to do that.
Surely the problem is that we've gated help,
any sort of assistance,
behind a diagnosis
that people have to fight for,
for years on a waiting list,
just to be believed.
Frith looking at that mess describes the problem is the label.
I think the problem is the gate.
And so finally,
the word,
the one that I asked you to hold right at the start.
She said it was dangerous to cling to the autism label.
I do not think the dangerous thing is a late diagnosed woman
finding a word that explains her life.
I think the most dangerous thing is telling a quarter
of a million people stuck on a waiting list
that the word they are waiting for is meaningless
in the middle of a fight over whether to fund them at all.
The danger was never the label,
the danger is the withdrawal.
The danger is careless statements from ivory towers.
That is why the title of this video says what it says,
not because Uta is a bad person,
she clearly is not.
She's a brilliant one who did real good
and has talked herself into a genuinely dangerous place.
Being brilliant has never once stopped anyone
from being dangerously wrong.
If this video helps something stick into place,
stim the subscribe button.
It genuinely tells the algorithm
that this is worth someone's time
and to show it to the next person who needs it
and leave a comment telling me where you landed,
even if you think I got something wrong,
especially if you think I got something wrong
because that's half the fun.
And if you want to see where all this started,
I made an earlier video on her first round of comments,
the calmer one before she doubled down.
And that one is the one to watch next,
it's right.
Also in the comments is a deep dive
on the four subtype studies,
the parent reported safari funded ones we talked about.
That's right here too.
And it goes a lot deeper into the eugenics adjacent
end of the swimming pool than I did today.
And if you'd like more of this
and you would like to help me to keep making it,
I do have a Patreon and I wanna be really clear
about how it works because I never want anyone
to feel pressured.
Everything I make for YouTube stays free always.
What it actually pays for is all the unglamorous stuff,
the software,
the Yorkshire tea,
the music,
the footage,
the licenses,
cat biscuits,
the servers,
and genuinely enormous numbers of hours
that a video like this one swallows.
A long form essay like today is the better part
of a working week and I've been doing most of the writing
for this video with one working eye.
if you've missed it.
I made two short films on the process
of my cataract diagnosis and surgery
from my perspective as an autistic adult.
And there is a lot in there,
if that interests you,
those videos are in the description below.
The lowest tier is the kitten tier at four US dollars a month.
And the main thing it gets you,
other than a selection of the finest exclusive videos,
is a really,
it's not really a perk,
it's a place.
It's a key to the scratching post discord community.
It is, and I say this as a man with 16 cats
and a professional interest in cozy.
One of the kindest,
warmest corners of the internet I know.
Hundreds of neurodivergence, and some neurotypicals too,
being gentle with each other,
sharing memes,
sharing wins,
sharing the odd bad day.
Plus around four exclusive members only live streams a month.
There is no expectation that you talk,
plenty of people just lurk and warm their hands on it,
and that is completely fine.
Your name goes into the credits
at the end of these videos,
at the food bowl,
sorry, at the credits,
unless you would rather I didn't,
in which case just tell me.
And if Patreon is not in the budget this month or any month,
please do not give it a second thought.
The single best way to support me costs nothing at all.
Watch the videos I'll be uploading to YouTube.
Stim the subscribe button and leave a comment
and tap the like.
That helps more than you know and is free.
Please be kind to the late identified autistic person
in your life,
even if that person is you,
especially if that person is you.
And I'll see you in the next one.
Bye for now.
Can't find what you're looking for?
Get subtitles in any language from opensubtitles.com, and translate them here.