AIght_
ToolsLearnFieldsUniverseSignalHumanAbout
Take the quiz
← All fields

Field guide

Psychology & Mental Health

AI is scaling access to evidence-based support through always-available chatbots and predictive insights, while augmenting therapists with data-driven session tools.

Medium
Start your path →See your personal disruption score in 2 minutes

What's changing

01

AI chatbots deliver CBT or DBT-based interventions 24/7, providing immediate coping strategies and mood tracking for users waiting for human care.

02

Ambient AI note-taking and clinical decision support tools analyze session transcripts to surface patterns, suggest treatment adjustments, and reduce administrative burden.

03

Predictive models from wearable and self-report data flag early signs of relapse or crisis, enabling proactive outreach by clinicians.

A model trained on therapy transcripts is impressive on transcripts. The hard part — the things that don't show up in transcripts — is exactly what therapy is for.

A path through the universe

How to actually learn AI for Psychology & Mental Health.

Two tracks. Pick your depth. The left one gets you fluent for conversations and tool choices. The right one is what you read when you actually want to know how it works.

Intuitions

No math required.

  1. 01Hallucination & GroundingWhy AI models confidently make things up — and what you can actually do about it8 min read
  2. 02AI Safety & AlignmentThe problem of building AI that reliably does what you actually wanted — not what you literally asked for11 min read
  3. 03Context WindowsWhat the model can see right now — and why the edges matter6 min read
  4. 04Prompt EngineeringThe craft of talking to a model that will take you exactly as literally as it decides to7 min read
  5. 05In-Context LearningHow models 'learn' from examples in the prompt — without changing a single weight.6 min read

Goes deeper

Under the hood.

  1. 01Retrieval-Augmented GenerationHow AI learned to look things up before opening its mouth8 min read
  2. 02Fine-TuningTeaching a model new habits, not new knowledge8 min read
  3. 03RLHFHumans rate, model learns, weird things happen — the post-training that made models pleasant to talk to.7 min read
  4. 04AI AgentsWhen AI stops answering and starts doing — and then, very often, hits a wall9 min read
  5. 05Constitutional AIWhen the model judges itself — Anthropic's bet on alignment without exhausting the rater pool.6 min read

AI impact spectrum

Automated

  • Intake documentation
  • Risk screening tools
  • Session note drafts

Augmented

  • Treatment plan suggestions
  • Crisis resource routing
  • Progress pattern detection

Growing

  • Therapeutic relationship
  • Trauma-informed care
  • Human empathy work

Roles at risk

Intake coordinator

Routine psychoeducation provider

Basic skills coaching support worker

Roles growing

AI-augmented therapist

Mental health technology integrator

Crisis triage specialist

Therapeutic AI ethics reviewer

Triage and screening compress. Long-term therapeutic relationships don't, and won't, for reasons that aren't technical.

What to actually do

Mental health professionals should integrate AI chatbots as adjunct tools for clients between sessions, using session-recording AI like Upheal to generate draft notes for review rather than starting from scratch. Review predictive risk flags from client wearables weekly and follow up personally, while maintaining strict boundaries on AI use in crisis situations. Complete ethics training on AI disclosure and bias, document all AI-assisted interventions, and use freed administrative time for deeper therapeutic alliance-building—treating AI as a scalable support layer that extends reach without replacing the human relationship essential to healing.

If you're a clinician, audit a Woebot-style flow as if you were the client. The things that work and the things that grate are both diagnostic.

Sources

  1. [1]Heinz et al., Randomized trial of a therapeutic chatbot for depression (2025)
  2. [2]APA, Artificial intelligence in psychology — Policy brief (2024)
Medium

Ethical and HIPAA concerns require careful tool selection and training, but compliant platforms exist and integrate into practice management systems.

Personalize this

How disrupted are you, really?

Three questions. An honest score tailored to your specific role.

Take the quiz →

Tools to know

Woebot

CBT-based conversational agent that delivers structured therapeutic exercises

Upheal

AI documentation and session insights for therapists, generating notes and progress summaries

Limbic Care / Clare&me

24/7 AI companions offering empathetic mental health support and triage

Concepts to understand

Conversational agents for therapeutic dialoguePredictive analytics for risk and relapseAmbient clinical documentation AI

Get your personal disruption score

Based on your specific role within Psychology & Mental Health

Run AI impact quiz →Explore other fields