Observations May-June, 2025
Autism and AI and Psychiatry, Oh My
Although Substack looks like a blog and quacks like a blog, it’s technically an e-mail newsletter.
This particular newsletter contains 2 months of short-form opinions, graphics, and commentary on science and medical news, selected and edited from Substack’s X-lookalike, Notes.
New Blog Posts
Free post: One of the most misunderstood things a neurodivergent person can say is “I can’t.” Here’s what it really means.
"I Can't" Means "Not Now"
Neurodivergent people are often misunderstood, and even mistreated, when saying we can’t do something. The internet teems with stories about it. Bosses can treat us as an uncooperative, uninterested employee. Friends and family may dispute whether we
Paid subscribers got 2 experimental updates:
Why Language Makes it Hard to Write About Neurodivergent Experiences
Writing about neurodivergent experiences is hard, because the nature of language works against me.
and, as promised,
Science and Health News
The US Health Care System Now Treats Mostly Chronic Conditions
According to the United States government Center for Disease Control (CDC): “Ninety percent of the nation's $4.5 trillion in annual health care expenditures are for people with chronic and mental health conditions.”
Yet, our system of medical care, organized around siloed specialties, is designed for those 10% acute emergency situations.
Historically, that’s where the US health system excels, to the point where even people from Canada & other countries have come for care.
Excited to read this paper from academia.edu about how to communicate with autistic people in medical care. Anecdotally, I hear a lot of stories from autistic people about difficulty communicating with doctors.
Full disclosure: I discovered this paper because it cites an article I coauthored. 😆
Guidelines for Randomized Controlled Trials Updates, Now Includes Open Science
When scientific researchers publish articles, they follow a set of guidelines for the type of study they’re doing: meta-analyses, randomized controlled trials, case studies, and more.
The guidelines for randomized controlled trials, CONSORT, just updated.
“The CONSORT 2025 statement consists of a 30-item checklist of essential items that should be included when reporting the results of a randomized trial” plus “a diagram for documenting the flow of participants through the trial." By "flow of participants,” they mean how many were recruited, how many were excluded from the study, and how many dropped out at each step.
Substantive changes include a new section on open science!
(Learned this news from Crystal Herron’s newsletter -- thanks!)
U.S. Defense Department Considering Replacing Obsolete Technology with AI
Not to sound like a crank, but is anyone else worried about giving AI control of the US’s nuclear weapons?
Because that’s in the new “Big Beautiful Bill”, in the part about modernizing the defense department. The law orders the department to replace current obsolete technology with AI.
(AI is made in our own image, and doesn’t have a conscience. What could go wrong?)
Opinions
If you know what the results of your investigation will be before you start, you may be doing research, but you’re not doing science.
Even if you’re trying to replicate a previous study: in that case, you don’t know whether you’ll get the same results or not.
How Psychiatry Works: Informed Trial and Error
My first neurologist, a kindly old man, forever shaped my understanding of brain-affecting medications. He said, roughly:
“We have a lot of theories about how migraines work, but we don’t know for sure. We know certain meds help, but we don’t know why. Treating them isn’t a science, it’s more like trial and error.”
Now, that was around 2010, before the newest batch of medications came out, including my current lifesaver Nurtec. We now know more about migraines and have better treatments. Yet, I continue to see evidence that prescribing can be more like an art or craft than a science.
(Especially in cases like mine, where a patient has multiple conditions. For example, it can be hard to treat depression without exacerbating anxiety, and vice versa).
In my experience, psychiatrists often follow this cycle:
1) Guess: they make an educated guess at an appropriate drug and dosage;
2) Test: they try it out;
3) Results: I report how I feel and they observe my demeanor;
4) Tinker: they adjust. Most often, they change the dose. But they may also stop or switch a current prescription, or add an “adjunct."
In short, medicine can be very different than hard science and math. As Aristotle might say, that doesn’t make it worthless. Still, psychiatry aspires to be a science, and to know things with certainty.
If patients expect that from their psychiatrist, they may be disappointed.
Psychiatrists need to be honest with patients about what they can offer.
Possibly Unpopular Opinion on Cognitive Behavioral Therapy (CBT):
Cognitive Behavior Therapy for depression and anxiety often assumes that your negative thoughts are illogical.
In fact, you often are reasoning logically, but from incorrect premises: assumptions about how the world works, often learned in childhood, that no longer apply.
Instead of playing the philosopher and asking, “is this thought logical?," it would be better to play the experimental scientist and ask, “is this thought correct in this case?"
You’ll probably argue with yourself less and learn more about the world.
ADHD is Starting Over…and Over…
You know how on January 1, you leap out of bed, eager to start on your New Year's resolutions, but by the end of February you realize you’ve stopped following them, and wonder what happened?
That’s what making a habit feels like to a person with ADHD, every time, all year.
You constantly get derailed when “life happens”, and then have to start your habits over again.
It’s a constant, exhausting emotional cycle, with energizing spikes and demoralizing lows.
Eventually, you might get sick of the roller coaster ride. Maybe you dread attempting a new habit at all, knowing it’ll kick off the cycle of disappointment. (If you procrastinate a lot, that could be a reason!).
It is Neither Possible nor Desirable to Find a Single Cause of Autism in Three Months
I really hope RFK Jr is joking when he sets a goal to “find the cause of autism” by September 2 -- that is, in 3 months.
Because that’s patently absurd.
Think about it: the world has been researching what causes autism since at least the 1990s, and poured big bucks into it since the 2000s (especially since Autism Speaks formed in 2007). In all 35+ years, after spending billions of dollars, we still have found no definitive cause.
So, of course it’s realistic to expect a breakthrough in the next 3 months. /Sarcasm
Leave aside the fact that research takes time. 3 months would be enough perhaps to design a study, recruit participants, and collect some data. It wouldn’t be enough to analyze and summarize the data.
Worse, the goal itself is a mirage. There is no One Cause of every case of Autism. That’s because there are many causes, and we’re still finding more. These include:
Many genetic mutations linked to specific syndromes, such as fragile X syndrome;
Hundreds of other genetic mutations, individually and in combinations;
Being born to very old or very young parents;
Being born premature, especially with birth complications;
Parents or siblings have diagnosed ASD;
Parents or siblings have high levels of ASD traits, even without a diagnosis;
Child has hearing problems or deafness, which lead to being deprived of language or communication stimulation at a young age;
Various pollutants and toxins that cause developmental delays in general. I haven’t followed the specific linkages, but am happy to look them up;
Extreme neglect, as in Romanian orphanages.
Actual diseases, including cancer, heart disease, and dementia, have multiple causes that involve both genes and the environment. Why should that not also be true of autism?
Someone this ignorant should not be in a position to set national health policy.
Thimerosal Timeline
Thimerosal is a compound once put in vaccines to prevent growth of viruses and bacteria. This chemical caused concern in the 1990s because it contains a form of mercury called “ethylmercury.” (Note: This is not the same type of mercury found in fish, which causes toxicity if consumed in high quantities—that’s methylmercury, with an m). At the time, people feared that vaccines in early childhood could cause autism or other developmental delays.
Some people still worry about this, even though thimerosal was removed from vaccines in the US. In 2001.
Thus, even if one believes that groupthink on this specific question somehow affected countless medical, scientific, and government organizations around the world for 30+ years, it’s a non issue.
And yes, I do mean there is consensus worldwide. For example:
August 2023: Another study looks for a link between autism incidence and the use of thimerosal-containing vaccines. The study does not find a link between thimerosal-containing vaccines and autism in Denmark and Sweden, where autism rates continued to increase although thimerosal was removed from vaccines in 1992.
February 1, 2009. Results of an American Academy of Pediatrics study found that immunization in infancy with thimerosal-containing vaccines does not decrease neuropsychological performance later in childhood. (That is, it did not reduce various cognitive skills, including executive functioning, working memory, & IQ).
(Moving from science to policy, is it a good use of time and money for an administration that claims to be concerned about the deficit, to investigate a threat that hasn’t even potentially existed for over 20 years?).
Key events:
The last remaining US vaccines containing thimerosal expired.
***2001
Except for influenza (flu), thimerosal was removed from or reduced in all vaccines routinely recommended for children 6 years of age and under manufactured for the U.S. market.
1999
The FDA examined the use of thimerosal in childhood vaccines and finds no evidence of harm. However, as a precautionary measure, it recommended removing thimerosal from vaccines routinely given to infants.
The CDC stated that vaccine manufacturers, the FDA, and other agencies were working together to reduce the amount of thimerosal in vaccines, or to replace them with thimerosal-free vaccines, as soon as possible.
This 2007 op ed in the New England Journal of Medicine explains that, when the US government admitted to lacking knowledge about vaccine effects, the public interpreted that as positive knowledge of harm.
The language used in medical research can be off-putting. Take “treatment-resistant depression”, which persists after trying at least 2 antidepressants. Researchers sometimes describe “patients who received and failed to respond to…”
Seems to me, the patients didn’t fail. The treatments failed them.
Writing about neurodivergent experiences is hard, because the nature of language works against me.
With all respect to Tolstoy… I’m pretty sure the happiest moment of my life won’t come while I’m washing dishes.
Will AI be Granted Rights?
How will people react if AI ever gets as good as humans at imitating expressions of human reasoning, empathy, arguing using evidence, or creativity? If AI becomes indistinguishable from us in its output, and its “thinking” is unknowable, we’ll inevitably ask, should AI be seen & treated as persons?
I fear that, rather than elevate AI and give it rights, we will denigrate humans & treat them like empty machines.
Is it a “Hallucination” or “Misinformation?” Depends if the Speaker is an AI or a Human With Unpopular Beliefs
Have you ever noticed how differently the media describes humans and AI when they say incorrect things?
When AI “says" things that are untrue, the media calls it “hallucinations". That implies AI can sincerely believes something untrue, for reasons beyond its control and not its fault.
When humans say something untrue, the media calls it “misinformation.” That implies that the person doesn’t believe what they say and is deliberately lying, likely for political reasons. Historically, it also implies the speaker should be silenced.
Why do we give AI the benefit of the doubt, but not our fellow human beings?
When I look for fiction featuring neurodivergent characters, I want, not to be “represented," but to be seen.
Research Tools for Everyone
Have you ever wished doctors would translate medical jargon into everyday language? Here’s a website that does that for you: https://www.justplainclear.com/en
Weird Facts
Apparently the Germans had a name for people with ADHD traits:
“The family name Lottering is of German origin, deriving from the word "lotter," which means someone who is careless or thoughtless. In European origin, the name likely referred to someone who was seen as lazy or lacking attention to detail.” -- https://crestsandarms.com/pages/lottering-family-crest-coat-of-arms
That’s all for now!
Do you have any comments, or news or ideas of your own to share? Hit the comment button below.