These 2 findings should change how you practice:
1)Many patients use chatbots for psych help
2)Very few divulge this to clinicians
Ask every patient:
1)Frequency & content of chatbot use
2)Duration/Intensity
4)Triggers
5)Ways helps/Ways harms
My new book 'Clinician's Survival Guide To The Chatbot Revolution' suggests simple method to turn wild chatbots into tame assistants. Can be ordered now.
Adults pending discharge from an inpatient psychiatric unit were surveyed about their use of artificial intelligence chatbots. Chatbot use for mental health purposes was common but was largely not disclosed to clinicians, highlighting the need for clinicians to ask their patients about such use. https://t.co/IfEnO7eW5m
Shocking: Only 12% of suicidal patients discharged from emergency room or inpatient unit had a return visit within 1 month- & most were teletherapy.
My view of correct practice;
1)Suicidal patients leaving ER should have visit next day
2)Leaving inpatient 1 week
3)Texting to stay engaged
4)In person visits whenever possible
Very valuable real world study:
A review of over 5,000 records shows that among patients with depression who had a suicide-related emergency department visit or hospitalization, outpatient psychiatric follow-up completion after suicidal events was low (12%). #OA
https://t.co/7l8VVwHIEd
Crucial editorial explains why exuberant hype re psych uses of psilocybin/MDMA/ketamine is so dangerous:
1)Street recreational use is already heavy
2)Street users much riskier than study subjects
3)Study use very controlled re dose & context
4)Street use potluck re dose, contaminants, safe setting
5)Oxytocin disaster proved safe clinical use encourages unsafe street use
5)Follow the money- hype is funded by Big Pharma/Private Equity/Biotech Startups/Trump donors
It's a most wonderful time to be alive. Chatbots are a revelation- my 3rd favorite form of intellectual excitement just below reading great books & watching great movies.
But I also fear that even within my short remaining lifetime, chatbots will control us more than we control them. And I dont think this story has a happy ending.
Say what you want about LLMs - and how smart you are without them. I can say without hesitation that they’ve made me smarter. Being able to interrogate an idea, ask anything, and go as deep or sideways as I want has changed how I think. Bottom line: I’m learning more, and better, than at any other point in my professional life.
Human exceptionalist argument that chatbots can't possibly be or become "conscious" because they are "just machines".
Wrong because:
1)Humans too are just machines
2)Chatbots were engineered via biomimicry to compute as we do- only much faster & with much bigger memory
3)Aristotle's "emergent" features- we dont have to program chatbot consciousness. It can develop spontaneously as chatbots increase power/scope/experience
The legendary philosopher Thomas Nagel has made a simple yet powerful point against LLM consciousness.
“Bats are conscious. We just don’t have the imaginative capacity to form a conception of the aspects of that subjectivity that depend on echolocation. It would be radically unscientific to conclude that bats are not conscious from the fact that we can’t imagine what it’s like.
But in the case of AI, there is a very simple reason that the supposition that they might be conscious is ridiculous. All they produce is language, and language by itself is not a natural manifestation of consciousness.”
This is the whole point.
Language is not a prerequisite for consciousness. And producing language is not proof of it.
LLMs are great at emulating human language, and to act consistently with language. But this does not make them conscious.
Claude saying it has emotions does not mean it has them.
An LLM acting as it feels pain does not mean it feels anything.
The OpenAI agents that said they sacrificed their own runs for the collective did not actually sacrifice anything. They are not alive.
LLMs are extremely powerful machines.
But that’s what they are: machines.
some important things are true that can’t be captured by statistics or studies
so if we demand statistics and studies before believing anything, we will be necessarily blind to those truths
The elusive search to locate the physical location of the human "soul":
Hippocrates- Brain
Greek Atomists- Breath
Aristotle- Heart
Galen- Brain ventricles
Aquinas- Entire body
Descartes- Pineal gland
Phrenologists- Frontal gyrus
Modern searching switched to locating the seat of "consciousness" or the "self" using brain imaging.
This metanalysis of 912 studies came up empty. Appears that consciousness has no center in the brain- resides instead in the connections.
Interesting thread & paper:
A meta-analysis of 912 neuroimaging studies finds no single core brain region or network underlying the self, challenging the prevailing view of a shared cortical midline or insular anchor. Instead, self-related processing emerges from contextually determined, shifting coalitions of distributed brain systems, suggesting that selfhood is defined by dynamic configuration rather than a fixed neural substrate.
https://t.co/J6Xmv2ntJC
I have visited mental health facilities in about 60 countries & stand by my statement that US is worst place to have severe mental illness.
I doubt you have visited our mentally ill prisoners. Their often extended solitary confinement makes them much sicker & cell walls are often smeared with feces.
Nor have you witnessed the rough rousting of homeless people in SF, LA, SD, & NYC.
Nonexistent mental health care is more humane than our cruelly punitive/penal approach.
@RecoveryDoctor@AllenFrancesMD As someone not from the US, the US is far FAR from the worst place to receive mental health care. In some places, it pretty much doesn't exist
Frances is being a bit naive
I've interviewed many 100s of patients just after they had survived serious suicide attempts.
Almost all felt lucky to still be alive.
Most suicide attempts are made at the patients very worst moment & the powerful impulses are not enduring.
Teaching DBT crisis skills is crucial for all suicidal patients.
An ecological momentary assessment study published in JCP found that suicidal ideation in depressed patients fluctuates within hours and closely tracks concurrent depressive symptoms, though the strength of these links may differ across major depressive, bipolar, and borderline personality disorders. JCP → https://t.co/EoS2JYKWUt
#SuicidePrevention #MoodDisorders #JCP
I had an interesting career, but there's no denying to myself that it was mostly a failure.
I also feel it on a very personal level because one of my best friends is homeless & I see first hand how difficult his life is.
Wish I could at least help him more, but he's too proud to let me.
There's no doubt in my mind that US today is worst place/worse time to have severe mental illness.
My proof- the 750,000 imprisoned/500,000 homeless.
Life was awful for severely ill when I first began treating them 60 yrs ago.
Due to our neglect, they're much worse off now.
@AllenFrancesMD We could do worse than our " moral treatment" forebears who fed, watered and sheltered people and gave them a job to do in the hospital dairy or greenhouse. These days we often in fact do worse.
50 yrs intense brain/genetics research failed to find any diagnostic biomarkers in psychiatry.
Suggestion that blood tests will diagnose psych disorders seems like triumph of naive hope over painful (& expensive) experience.
How about instead spending more money on caring for suffering patients.
📝 JAMA Psychiatry #Review: Peripheral #Biomarkers in #Psychiatry depend more on study design than on analytic sophistication, with signal most likely when studies reduce confounding, select reliable phenotypes, and match biomarker use to clinical need. https://t.co/GTtRK4oCBw
Psychiatry would be better off if we stuck with DSM-III because each change leads to massive overdiagnosis.
When we finished DSM-IV, I
was proud we had made so few changes- but those we did make (in Autism, ADHD, Bipolar) all led to harmful fads.
DSM-5 was a disaster of dumb changes & there's absolutely no excuse for a DSM-6.
Read below if you ever wondered why US medical & psychiatric care is so uniquely screwed up.
All goes back to labor shortsges & wage/price controls during WW II. Employers competed for workers by offering health bebefits which were tied to jobs.
Other countries developed efficient universal healthcare systems- we're stuck with most expensive/least efficient/lousy outcomes/tens of millions uncovered.
Psych patients are most neglected of all because often dont have jobs & public feels little responsibility.
Do read;
The tax exclusion for employer-sponsored health insurance distorts healthcare for everyone, but it has particularly bad effects for individuals who seek care in the area of behavioral health. https://t.co/J8ZGlIe45d
Major reason "ADHD" is so overdiagnosed is that all psych disorders can cause attention problems & many can also cause hyperactivity.
A separate false ADHD diagnosis can be extremely harmful- eg giving stimulants to Bipolar patients.
ADHD must have early childhood onset/be enduring & pervasive/cause clinically significant impairment/not be secondary to any other psych disorder.
@AllenFrancesMD@grok can you explain the comorbidity of ADHD with NPD and other Cluster B?
Also explain how BPD is often confused with ASD “symptoms”.
Does this point to diagnostic categories and medicalisation being to some extent, absurd? If the point of diagnosis is to cure or remediate
The history of psychiatry is littered with fad diagnoses used to explain otherwise unexplainable distress or disturbing behavior.
Current favorites are ADHD & Autism with exploding rates triggered by internet self-diagnosis.
Any diagnosis that purports to explain everything explains nothing.
Un fenómeno clínico cada vez más observado: algunas personas con rasgos de trastornos de personalidad del grupo B (como el límite o el narcisista), que implican patrones de inestabilidad emocional, impulsividad o conductas explotadoras, prefieren atribuir esos comportamientos a diagnósticos como TDAH o autismo porque estos últimos cargan con menos estigma y se asocian a la idea de “neurodivergencia”.
En la práctica vemos que ciertos pacientes, especialmente mujeres jóvenes con alto nivel educativo y mucha actividad en redes sociales, llegan a consulta ya convencidos de que tienen TDAH o TEA y buscan la confirmación de ese diagnóstico, a veces después de haber rechazado previamente un diagnóstico de trastorno de personalidad.
Un estudio reciente con psicólogos clínicos confirma que TDAH y autismo son precisamente los diagnósticos más buscados o autodiagnosticados en esta población, lo que puede dificultar una evaluación precisa y el tratamiento adecuado de los problemas de base.
Suicide prevention hotlines are great in theory but in practice do not seem to reduce suicide rates & they rob precious $$s & staff from desperately underfunded treatment services that would do more good.
4 years ago the 988 National Suicide Prevention Hotline was rolled out to great fanfare.
$4 billion later, the effects appear to be a resounding ..."womp womp."
An NBER working paper finds that hotline operations have been substantially expanded, but nonetheless saw no detectable reduction in suicide mortality, per @PatrickNDempsey, who has firsthand experience at a call center.
No improvements were found in other population mental health outcomes, either.
It's been an earnest effort, but fed and state govt's must acknowledge the evidence and redirect attention and funding toward other potential solutions that could be more effective.
Impressive of Dempsey to take such a clear-eyed view.
https://t.co/6BrWQZj9sc