Tom Marshall

One First Priority — A Translational Model of Decision Making in a Therapeutic Context

Meet Tom Marshall

Hi. I'm Tom. Welcome to my temporary page. It's not pretty, but it'll give you a taste of my work and ideas to get you oriented while my permanent site is under construction.I'm a neurodivergent physicist with a background in applied math and software engineering, which gives me an excellent toolkit for exploring other fields.The one that's most important to me right now is how people make decisions. But since pretty much everything that you do every day involves making many decisions, large and small, decision-making turns out to be a very important life skill.This interest inspired a deep dive into neuropsychology and evolutionary history — which showed me an effective, practical path to help people understand themselves.That's the main mission of this site, but I also have many other interests that I'm eager to share with you. For me they're all interrelated, but I can see that some of you will want to prune the tree a little bit.This page will let you get a taste of all of the projects — and learn more about the ones that interest you most!

My ProjectsOFP The Book
I'm writing a book that might have the subtitle "A Practical Guide to Your Mind". It's rooted in decision making, and it'll give you practical tools for self-understanding and for making your life better on your terms, whatever that means to you. I'm still working on it, but I'll be sharing some details and snippets here.
OFP and Neurotype
It's easy to see the differences in people's cognitive and emotional styles, but what struck me — after a lot of exploration — was that the similarities are much more central and fundamental to who we are than the differences. I say this as a scientist; I'm not making any value judgements.
This line of thinking got me interested in the more academic topic of what this neurotype that we all talk about actually is. Turns out "neurotype" isn't a well-defined technical term. I'm trying to change that — for very practical reasons as well as for academic curiosity.From my perspective, everyone has a neurotype, which for now we can think of as a whole bunch of "sensitivity setting knobs" that adjust the functioning of our brains — but everyone's brain is more the same than different.Think of a stereo. You might like more treble brightness at a lower overall volume, where I might prefer a booming base that supports the melody. But the stereo hardware is the same, and all stereos can play rock and rap and classical and jazz, even if your stereo happens never to play baroque and mine never plays opera — but they all can.This academic work may not interest you, but it has informed my understanding of my own mind, and maybe I can use it to help you understand yours.LLMs (What Some Call "AI")
Also, since I use LLMs such as Claude, ChatGPT, and Gemini as research tools, I've explored those tools rather deeply. You may note that I didn't call them "AI" because I don't think that those computer programs are intelligent.
But we don't have a good definition of "intelligent", so maybe they do meet your definition. My point is that we don't really know why they work as well as they do. We do know how they work, but not why. And I prefer not to project too far into the unknown unless I'm calling it a conjecture.Physics
And then there's physics. You won't find a thorough physics tutorial here, but I will share some thoughts and conjectures that I think are both technically interesting and useful for what I call "cultural literacy" in the sciences.
Stuff About Me
Also, since I only recently discovered my neurodivergence (there's a little story on that in the Bio section), that got me interested in how different people approach different situations.
Onward!
The next few sections give brief overviews of my areas of interest. Each one has links to documents that'll give you more information on specific topics.

Apologetic NoteMany of the documents that I've splatted onto this temporary site are written in my unfortunately dense academic voice. I promise that the book — and any newer documentation — will always be in much more accessible tones, even when the subject matter gets a bit technical. "If you can't explain something simply then you don't understand it well enough."

I: One First Priority — A Practical Guide to Decision MakingOne First Priority (OFP) is, under the covers, a translational model of psychotherapy that addresses the real world directly. But since I'm not a licensed therapist, I focus on building and sharing practical tools that can make your life better no matter where you're starting from. The key skills are self-knowledge (which is a cognitive skill), self-awareness (which is an emotional skill), and the oftentimes daunting task of finding out what it is that you really want — and prioritizing them into a "portfolio of wants" that you can then make plans to achieve.OFP The Book
I'm in the process of writing a book on this work. It's focused on very practical skills that you can put into practice in small steps — although there's a lot to understand to get the really best results. That's why it'll take a whole book.
Meanwhile, here are some documents that you can get started on. Remember that this is a temporary page where we can get to know each other a little, but I understand if things aren't clear enough for you now and you want to come back later.Some OFP Documents
Here’s a high-level overview to give you a small taste of OFP.
Here’s a more concrete framing of OFP that was generated for me by ChatGPT — don’t barf, the next document shows you my words that it used. The whole process was rather enlightening!And here’s the work that went into generating the compact framing — you see, it really is my work product!Here’s a slightly whimsical essay that captures the connection between our behavior and our biology. For me it’s very meaningful. Perhaps if you think on it as a Zen koan it will become clearer.OFP is about decision-making, which we like to think of as a very “rational” process, but when you consider the whole decision process including the choice of problem to be decided in the first place, the assumptions made before starting analysis, the rational analysis itself, and the implementation of the decision, it turns out that the whole process is not “rational” — and understanding this is very important for making decisions that actually meet your needs and make you happier.Emotional Binding (EB) is when something becomes real enough to act on with little or no friction. EB often happens spontaneously — we’re emotionally attached to many of our beliefs that guide us through our day — but it’s also a skill that can be learned to help you towards goals that you believe in or want cognitively, but “just can’t get there” — like diet or exercise.Here’s a more academic explanation of Emotional Binding.

II: NeurotypesOFP is a biologically based framework that's significantly informed by the nature of neurotypes. Resent research in molecular genetics strongly suggests that "the autism spectrum" is just the tail of what should be seen as "the neurotype spectrum", which includes the entire population. At present, neurotypes are best understood as an array of "sensitivity settings" of various neurological receptor and filter systems.OFP therefore looks for friction between an individual's neurotype and environment and seeks a mutual accommodation. OFP doesn't equate normative behavior with healthy behavior. One can, through personal agency, choose to accommodate the environment when the goal is worth the cost, but that accommodation per se is not a goal. OFP sometimes uses the language of "donning a hazmat suit to face a hostile environment because a task provides sufficient value." This again emphasizes agency.This masking, whether deliberate or not, and whether entered into freely or seen as a drastic, necessary survival tool is not limited to the autistic or other neurodivergent communities. Most neurotypes frequently find themselves at odds with societal normative-behavior situations, especially institutional ones such as the workplace.OFP expands Damian Milton's concept of double empathy, which in OFP is simply called symmetry, to state that there are a priori no preferred neurotypes. Although some neurotypes do cross the boundary into pathologies, OFP regards the demarcation line as being environmentally and personally influenced, that is to say, not purely objective.What Do I Mean By "Neurotype"?

  • Rather than viewing neurodivergence as a categorical departure from a uniform norm, the neurotype framework states that all individuals occupy coordinate location in a continuous, multidimensional neuromodulatory manifold. In this perspective any given neurotype has a scalar distance from the population mean. That distance says little about the individual coordinates of the individual.

  • I don't yet have a full mathematical model, but one toy model has four axes: sensory gating, synaptic filtration, receptor sensitivity, and neuroplastic adaptation rate. If you don't know what these axes mean right now, don't worry. It's just a toy model anyway

  • This model rejects a binary taxonomy of human brains. Instead, it positions all individuals on a shared continuum of signal-to-noise modulation. Psychotherapy, therefore, acts as a regulatory tool to help the patient optimize processing strategies based on their specific location within this universal distribution.

  • The model treats individual cognitive- and emotional-style variations as a spectrum of different settings within the nervous system (including the peripheral and enteric nervous systems, which play major roles in the total neurotype). What are called "clinical presentations" are frequently the behavioral manifestations of non-normative configurations or sensitivity settings of standard, universal functional units within the nervous system.

On the Art and Science of DiagnosisData suggests that only 5-15% of the general population who present themselves for therapy, voluntarily or otherwise, are affirmatively diagnosed to have an underlying neurological or other medical condition. There are hints from less formal therapeutic guidance that there is an even larger percentage of the general population who have some form of psychological distress resulting from the interactions of their environment with their specific neurotype and neural history.Clearly trauma, psychological, physical, or otherwise, can lead to sufficient dysregulation that conventional psychotherapeutic modalities will provide significant therapeutic efficacy and improvements as reported by patients in quality-of-life indices. It is less clear that this course of treatment is uniquely optimal.Some Documents
Here’s an essay showing the relevance of neurotype to the whole concept of clinical psychology: What Autism and Genetics Have to Teach Us About Psychotherapy
As the above essay shows, I’m deeply interested in the diagnostic as well as the treatment aspects of psychotherapy. Here’s something I wrote in response to recent articles discussing improvements to the Diagnostic and Statistical Manual of Mental Disorders (DSM). I think that many, or even most presentations that are diagnosed as “disorders” are perfectly healthy expressions for that individual. I’ll get into that more deeply in the next few documents. For now, here’s my commentary on thoughts on the DSM: Neurotype-Aware Contextualized DiagnosisHere’s a series of documents that show my thinking on building a clinically useful, mathematically rigorous mathematical model of the general concept of neurotype, as discussed above. They overlap a bit, so you might want to start with the last one and check the earlier ones as needed for more background.Towards A Clinically Useful Generalized Model of NeurotypeToward an Empirically Grounded Metric Space of NeurotypeTowards a Coupled Neurotype-Environment Geometry

III: Large Language Models (LLMs) — So-Called "AI"When it comes to LLMs, I think that there’s far too little data to support the notion that they’re “intelligent” by the definition that I think makes historical tangency with where the word came from in the first place — it’s always been applied to people and animals.I’ll say more on my definition of “intelligent” soon, but here are some observations on what LLMs actually are — and why they behave the way they do. With a plausible data science explanation and no embodied, self-benefiting behaviors, I still strongly resist any projection that these machines are intelligent. Note that I’m not asserting machines can’t ever be intelligent — although I’m skeptical — but I’m more confident that they’re not there yet.Here’s an accessible essay on my position: Miror Mirror in the Data CenterHere’s a report on work that I’ve been doing for over a year, trying to answer the question “What must be true about human cognition and use of language that such a conceptually simple, if large, algorithm as the LLM Transformer architecture can emulate so much of human behavior, including arrogant ‘strong and wrong’ answers?” I call this research the Inverse LLM Problem. The Inverse LLM Problem

IV: Physics

V: Additional Biographical TidbitsI've been working informally with the NYU Institute for the Study of Decision Making, attending their Neuroeconomics Colloquia and discussing ongoing research. I recently attended the NYC ACT Bootcamp.I have a very good friend who happens to be a clinical psychologist. When visiting her about three years ago, I read out loud a Facebook post about neurodivergence. I said to her, "But that could describe me!". She looked at me kindly and said, "Aw, Sweetie! You mean you don't know?" That was my diagnosis, and a lot of things make more sense now. I was aware of most of the neurologically odd trees, but I never saw the neurodivergent forest before then.