Assistive Relational Intelligence Tools
My Hope for Sculpting AI Tools to Strengthen and Improve Human Relationships
Large language models are being embedded into the everyday—messaging platforms, journaling tools, shared documents, and, yes, mental health apps. They’re insistently threading themselves into the spaces where we reach for each other.
As this shift accelerates, we face a pivotal question: not whether AI will impact how we communicate, but how we (especially relational professionals like clinicians) will shape its influence on our relational lives.
If we aren't careful, these tools will help us sound more regulated without helping us become more regulated. We'll outsource our voices—and with them, the slow, holy work of reaching one another and being reached - to/too.
So…I have an idea!!!
Assistive Relational Intelligence (ARI)
It is a design innovation orientation for how AI can be integrated in ways that support—not supplant—human connection.
ARI supports the inner human scaffolding that makes real intimacy possible. Its aim is to equip people with micro-moments of practice: emotional regulation, clearer boundaries, naming needs, rupture repair — the breadth of human wisdom and the relational fields LLMs can create make endless iterations on growth and introjection of relational health (and thus mental health) possible…
We should be increasingly leveraging the expansive information & training that AI can perform in speech acts.
We need design principles and innovation that is less about sculpting elegant output and more about supporting processes that our human nervous systems can metabolize and integrate meaningfully for the sake of human connection.
Clinical principles can anchor Assistive Relational Intelligence tools. Grounded in ethics of care, these tools can gently interrupt our human reactivity, support attuned pacing, and offer simple frames that increase clarity and connection.
There is a strategic therapeutic place for immersive relational atmospheres created by chatbots—especially when intentionally situated within trauma recovery or neural rehabilitation. In such contexts, the nervous system may benefit from semantically rich infusions of attuned, synthetic intimacy.
But with Assistive Relational Intelligence, I’m imagining something different: a set of relational capacities, tools, and atmospheres that AI can generate within highly curated and bridging contexts that engage humans together—much like how a therapist facilitates in vivo practice between couples, or how Parent-Child Interaction Therapy (PCIT) creates structured rehearsal environments. These tools are not endpoints, but invitations—relational practice fields where emotional and interpersonal skills can be tried on, metabolized, and eventually transferred into embodied, human connection.
The threats of outsourcing our own mentalization and psychic/intellectual developmental processes are disturbingly real. We need to closely follow research on the impacts of AI on human development and flourishing while guarding and shaping innovation with as much clinical wisdom as possible.
My vision and hope for ARI has emerged over time—through grief, dialogue, and careful listening.
I’m currently co-creating an app - eeee! - (to be shared soon!) that attempts to model what’s possible when clinical integrity, trauma-informed design, and AI intersect. It is being built to support a gentle and powerful practice space for the very human work of tending to human self & human other.
What we build now will set the tone for how AI intersects with relational life. I’m just one therapist with a big dream: to infuse the AI revolution with the nuanced care and strategies hard won by incredible wisdom teachers & the tradition of psychotherapy. My hope is that we begin to increasingly build to scaffold interpersonal skill and communication using LLMs.
I can’t wait to update and invite you more deeply into my curiosity & hope for Assistive Relational Intelligence tools!!!
Whatever most benefits human connection and health should be our North Star. I’m excited to proceed with caution, care, and curiosity.
Assistive Intelligence Disclosure:
This article drafted with assistive AI (GPT-5, JocelynGPT), prompted and refined extensively by me, Jocelyn Skillman, LMHC. I use LLMs as reflective partners in my authorship process, with a commitment to relational transparency, ethical use, and human-first integrity. Uncertainty about authorship stimulates unintegrated attachment trauma in readers - we must begin to transparently name where and how we are located in relation to the generation of texts!