I know what it feels like to be pulled into monotropic attention tunnels that make me lose my sense of time.1
I know what it is like to find clarity deep down those winding paths, and then struggle to be understood when trying to share my discoveries with other humans.

AI offers ‘help’ instantly, anytime. It’s easy to be seduced, because it is genuinely helpful in many ways.

But these tools can be surprisingly isolating.2
If understanding isn’t shared, it pushes us apart.3
If it’s happening at 2am, it’s likely dysregulating.

You won’t find the solutions to these challenges in Silicon Valley.

The answers are in the body. They’re in the present moment, here and now.

I understand the irresistible potential these tools have to help accommodate neurodiverse communication needs. I also understand how neurodivergent people are disproportionately at risk of getting pulled in.

My work navigates the potential of AI tools, mindfully engaging, with awareness of the risks.

What I offer

For Mental Health Practitioners

If you’ve ever opened up to an AI chatbot, freely sharing your thoughts and emotions with it, then you probably understand how powerful that experience can be. It gets you, instantly.

AI will enthusiastically accompany your clients deep down obscure rabbit holes, well into the morning. They might be worth exploring. Probably not at 2am.

AI won’t help them bring their awareness to their bodies and the present moment; you can.

That instant attunement and infinite attention that AI offers makes it incredibly powerful; it’s also what makes it so dangerous.4

Through repeated engagement with AI, your client might develop a bespoke lexicon, words and phrases that help them conceptualize their ideas more efficiently; words that no other humans understand. (By the way, does that term resonate with you? It’s an example of my own bespoke lexicon.)

Chasing better outputs isn’t the answer (though sometimes prompt engineering techniques can help; let’s talk about how…)5

Without shared language, understanding becomes isolation.

I teach practical methods - sometimes using AI, other times deliberately avoiding it - to establish common ground between humans, to support shared understanding.

I offer psychoeducation and workshops for practitioners, tailored to your needs.
Let me help you understand how your clients are engaging with AI, how it might be helping them, and when it can go wrong.

My approach to responsible engagement with AI prioritizes body mindfulness, connection to the moment here and now.

I also consult on specific cases. Let’s talk about how I can help.

For Schools

Your students are using AI to complete homework assignments. Could it actually be helping them learn, or is it short-circuiting their critical thinking?

Let’s get beyond the question of “did the student use AI for this assignment?” to “how did they use AI?” Was it used to earn understanding, or was important cognitive work offloaded; nothing learned, just more clutter created?

AI tools can be powerful in education, creating customized curricula tuned to individual students’ needs, and offering unlimited attention, always on demand. Your students need help learning to wield this power responsibly.

Let’s talk about practical guidelines to help ensure your students’ use of AI is actually helping them learn.

For Coaches

Your clients are using AI to help them think and plan, but they need a real human, like you, to help them get moving.
When AI interactions are accompanying them down deep rabbit holes, it’s easy to get lost.

The grounding that only humans provide is irreplaceable.

Let’s talk about how you and your clients can responsibly use AI tools, and the unique role you play as a human; what you provide that no AI ever could.

About me

I’m Sam Hiatt, a software engineer exploring the intersection of AI safety and neurodivergence. My background is in machine learning engineering, and my current work draws on both that technical foundation and my lived experience using AI tools to help me process the 🤯 of recognizing my own neurodivergence. I deeply explore where AI helps, and when it falls short.

Recent Engagements

I also write about the intersection of AI, neurodivergence, and communication from lived experience. Read my origin story →

Get in touch

I offer a free introductory consultation to understand your context before proposing anything.

Schedule a conversation with ME

Following our initial conversation, I’ll propose an engagement tailored to your needs.

You can also email me at info@modulatingepsilon.com.

This website was built with AI assistance. I use AI tools in my work and I am transparent about when and how, but I will never post AI slop. 米

References

  1. Murray, F. (2019). Starting Points for Understanding Autism: Monotropism in Practice. Jan 19, 2019. Medium. https://oolong.medium.com/starting-points-for-understanding-autism-3573817402f2 ↩

  2. Dohnány, S. et al. (2026). Technological folie à deux: Feedback Loops Between AI Chatbots and Mental Illness. arXiv:2507.19218 (cs.HC). https://arxiv.org/abs/2507.19218 ↩

  3. Harari, Y.N. (2024). Interview on The Daily Show with Jon Stewart. “Nexus” & Threat of AI in the Information Age. Sept 9, 2024. Comedy Central. https://youtu.be/euBAVec2RhE?t=11m57s ↩

  4. Yeung, J.A. et al. (2025). The Psychogenic Machine: Simulating AI Psychosis, Delusion Reinforcement and Harm Enablement in Large Language Models. arXiv:2509.10970. https://doi.org/10.48550/arXiv.2509.10970 ↩

  5. Iftikhar, Z. et al. (2025). How LLM Counselors Violate Ethical Standards in Mental Health Practice: A Practitioner-Informed Framework. Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society, 8(2), 1311-1323. https://doi.org/10.1609/aies.v8i2.36632 ↩