AI for Telehealth Therapy Without Mental Health Medications?

By David Stephen

Sedona,, AZ — There is a new book, Unshrunk: A Story of Psychiatric Treatment Resistance, [Published: March 18, 2025] with the summary, “At age fourteen, Laura Delano saw her first psychiatrist, who immediately diagnosed her with bipolar disorder and started her on a mood stabilizer and an antidepressant.  Delano’s initial diagnosis marked the beginning of a life-altering saga.

For the next thirteen years, she sought help from the best psychiatrists and hospitals in the country, accumulating a long list of diagnoses and a prescription cascade of nineteen drugs. After some resistance, Delano accepted her diagnosis and embraced the pharmaceutical regimen that she’d been told was necessary to manage her incurable, lifelong disease. But her symptoms only worsened.

Eventually doctors declared her condition so severe as to be “treatment resistant.”  After years of faithful psychiatric patienthood, Delano realized there was one thing she hadn’t tried—leaving behind the drugs and diagnoses. This decision would mean unlearning everything the experts had told her about herself and forging into the terrifying unknown of an unmedicated life.”

What, in the brain, becomes altered to result in a serious mental illness? Could AI map the mind and its alteration? If mind is assumed to be the same as mental, what is normal for the mind, to result in regular social and occupational functioning and what is abnormal to disrupt those?

In the evidence, since decades of neuroscience research, the closest cells to how the mind works are neurons. However, neurons [for all they are said to do] never function without their signals: electrical and chemical. When neurons are said to fire or be active, electrical signals are involved and chemical signals as well, in general.

So, if the way the mind [or mental] works is to be understood for normal or otherwise, it is possible to describe it with neurons plus their electrical and chemical signals.

Neurons are cells, like several others across the body. Their anatomy is near fixed, so it is unlikely that neurons can by themselves represent the memory of anything. This means that a neuron may not change shape to [re]present an emotion or have another for a feeling. Even in a cluster, it is unlikely that neurons would have shapes for every memory, given the large amount of memory a human may have, through the years—ranging from education, to places, people, objects, and so forth. The time it would take to change shape and the energy would be disruptive.

Simply, neurons, either singular or in clusters are unable to represent memory, emotion, feelings or regulation of internal senses, with shapes or by changing shapes. If this were the case, why would electrical signals have the ability to transport the shape of a cluster of neurons to another cluster, and what would the roles of chemical signals be?

To develop a concept of how the human mind works, the likeliest candidates [within the evidence in neuroscience] are the electrical and chemical signals. Not an individual electrical signal or a singular chemical signal, but electrical and chemical signals as sets or loops, available in clusters of neurons.

So, every function is approximately a particular assemblage, configuration or formation of electrical and chemical signals in a set. This means that to know what a table is, differently from a window or a fan, is to have electrical and chemical signals assembled in a particular way, in interaction. Electrical and chemical signals have to interact. Electrical signals, in a set, strike for the formation available at the set of chemical signals. The instance of the strike is when memory is produced, or emotion, or feelings, or regulation of internal organs, conceptually.

Also, sets of electrical signals and sets of chemical signals often have their states while interacting. These states grade or measure the extents to which they interact at that instance, for the function.

To develop a parallel architecture for mental disorders, for delivery by AI, as a source of therapy, it is possible to use sets of interacting electrical and chemical signals, with their grades, to place conditions, across severity. Hallucination could be the result of a problem of distribution of electrical signals across sets, for interactions to differentiate reality. There could also be lower than regular volumes for some sets of chemical signals, affecting their ability to be thick sets or thin sets. Thick sets hold whatever is common between two or more sets, while thin sets hold what is unique. An individual with several [non-collected] thin sets—which is what memories from birth often are, before they organize into thick sets—may have issues with delusions.

There can also be a problem of too much or too few intensity of electrical signals in some sets, or a lack of splits of electrical signals [with some going ahead of others, such that if the input matches distributions continue, and if not, the incoming electrical signals go to the right set of chemical signals, correcting the error. This explains the terms predictive coding, processing and prediction error.]

There could also be problems with a principal spot or principal measure where a set becomes dominant, stifling other sets from their regular [volume or intensity] minimums, like the heaviness in a major depression, that prevents doing anything else. There are sequences as well, old or new, and so forth.

Sequences for internal and external sensory inputs are the takeoff points of electrical signals, in a set, from a side of the set of chemical signals. If the path had been previously used, it is old, if not, it is new. Old sequence are useful for routines and procedures. New sequences for explorations and adventures. There are mental disorders that are problems of lack of prints of old sequences, making some experiences feel new, resulting in disturbances.

It is possible to describe all the conditions in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision [DSM-5-TR] with these, advancing the understanding of the mind, within the current evidence in neuroscience.

The difference between normal in the mind and otherwise could be an interaction, or a split, or a removal from the principal measure and so forth. Some medications may target certain neurotransmitters or their receptors, shaping their volume, or the way they accept intensities of electrical signals or prompt them. These might help in some cases, while side-effects may also follow.

There could be cases where certain adjustments [that induces interactions or grades of the mind] could lead to relief, and some where there could be need for medications. Progress in psychiatry [aside from antipsychiatry and critical psychiatry] is to explore electrical and chemical signals, for their possible role in mind order and disorder, which could be delivered by AI, as well.

There is a recent [March 20, 2025] report in Reuters, VA shake-up hits mental health services for US veterans, stating that, “The VA provides health care to 9.31 million U.S. veterans at hundreds of medical centers, clinics and nursing homes across the country. It also faces complex problems. The Trump administration plans additional cuts to the VA of more than 80,000 personnel, according to an internal memo obtained by Reuters. The agency has also announced it is phasing out telework. The VA employees – who include six mental health professionals and three people in leadership positions – described cancellations of some in-person and telehealth appointments; confusion over staffing of a crisis hot-line; and professionals conducting telehealth visits in makeshift meeting rooms inside VA buildings. The Veterans Health Administration, the branch of the VA that provides healthcare, has experienced severe staffing shortages since 2015, especially among mental health professionals, according to an OIG report last year. Veterans often benefit from specialized services to treat anxiety, trauma, depression and substance abuse. The proportion of veterans receiving mental health services rose to 31% in 2022 from 20% in 2007, according to the VA. Suicide among veterans is twice the rate of Americans overall.”