Tricia Friedman

My cast of bot characters to learn with

In the past year I’ve built over 100 different chatbots in order to learn when they are, and when they are not helpful in a learning journey.

My cast of bot characters

A timeline of purpose-built bots to try out.

Skim by theme below. Each card links directly to the bot.

A creator framework for learning bots

The four components of a learning bot

A learning bot is not only an information system. It is a designed emotional encounter with knowledge, challenge, perspective, and possibility.

DataDetermines what the bot knows. TechnologyDetermines what the bot can do. DesignDetermines how learning unfolds. StorytellingDetermines how the relationship feels.

Where does SEL belong? SEL is not a fifth technical component. It is the emotional and relational quality running through all four: how the learner experiences trust, control, safety, challenge, dignity, empathy, and agency.

DataThe knowledge, evidence, examples, voices, and boundaries that ground the bot.

Questions the creator must ask

  • What knowledge does this bot genuinely need to support its human capacity?
  • Which research, frameworks, examples, or professional practices ground it?
  • Whose experiences are represented—and whose may be absent?
  • Is the underlying language strengths-based or deficit-based?
  • What uncertainty, disagreement, or limitation must the bot disclose?
  • What sensitive information should it never request, retain, infer, or diagnose?
  • When must it say “I don’t know” or direct the learner to a human?
  • How will time-sensitive knowledge remain current?

SEL and emotional considerations

TrustRecognitionDignityVulnerabilityInvalidation

Will learners see their lives represented with care? Could the bot’s knowledge make someone feel reduced to a category, corrected about their own experience, or pressured to disclose something private?

Emotional aim: The learner should feel accurately informed without feeling defined by the data.

TechnologyThe capabilities, memory, privacy choices, limits, and failure modes of the system.

Questions the creator must ask

  • Which technological capability is necessary for the learning goal—and which is merely impressive?
  • Does the experience need memory, or would memory create unnecessary risk?
  • What happens when the system misunderstands emotion, identity, intent, or context?
  • How clearly will it communicate uncertainty and limitations?
  • Can the learner pause, revise, restart, slow down, or leave?
  • Could fluent language be mistaken for professional authority?
  • What privacy expectations must be visible before disclosure begins?
  • When should control return to a teacher, caregiver, expert, or learner?

SEL and emotional considerations

ControlPrivacyFrustrationDependenceUncertainty

Does the technology preserve learner control, or does it feel watchful, inescapable, or overly authoritative? How will the bot repair the interaction after misunderstanding someone?

Emotional aim: The learner should experience capability without surrendering autonomy.

DesignThe choreography of choice, practice, friction, feedback, reflection, and transfer.

Questions the creator must ask

  • What precise human capacity should grow through the interaction?
  • What will the learner actively practise rather than merely receive?
  • What does the first turn ask, and is that request emotionally proportionate?
  • Where does the learner have meaningful choices?
  • How quickly does difficulty or emotional intensity increase?
  • What makes the friction productive rather than punitive?
  • How does feedback protect agency instead of declaring success or failure?
  • What happens after confusion, resistance, error, or emotional overload?
  • What reflection makes learning visible, and what transfers beyond the bot?

SEL and emotional considerations

SafetyAgencyCompetenceDiscomfortShamePrideHope

Is the experience safe enough for a learner to risk being wrong? Does challenge stretch thinking without creating shame? Does the learner own the progress, or does the bot take credit?

Emotional aim: The learner should feel challenged, capable, and able to recover.

StorytellingThe role, voice, metaphor, point of view, narrative stakes, and learner identity.

Questions the creator must ask

  • Why is this bot a character rather than a conventional tool?
  • What does its role make emotionally and cognitively possible?
  • What power relationship does the character create?
  • Should its voice feel inviting, challenging, reassuring, playful, or authoritative?
  • Could the character be mistaken for a real person or professional?
  • Does the story encourage empathy without manipulating emotion?
  • Are any identities simplified, romanticized, or stereotyped?
  • Who does the learner get to become within the story?
  • How does the narrative close and return the learner to reality?

SEL and emotional considerations

CuriosityEmpathyBelongingDistanceIdentificationWonder

Can the fictional frame make difficult practice feel safer? Does identification widen empathy—or create over-attachment and misplaced trust? Is emotion invited honestly rather than engineered invisibly?

Emotional aim: The learner should feel drawn into the story without losing critical distance.

The emotional contract of a learning bot

Across all four components, the experience should support a deliberate emotional movement.

EnterSafe enough to begin.
ChooseFree enough to exercise agency.
LearnChallenged enough to grow.
RecoverSupported enough to repair.
LeaveEmpowered enough to act without the bot.
Creator planning checklistTen questions to answer before building.
Human capacity: What distinctly human ability should grow?
Data: What knowledge grounds the experience?
Technology: What capability is necessary—and what boundary protects the learner?
Design: What will the learner actively practise?
Storytelling: What role makes that practice meaningful?
Emotional entry: How might the learner feel when arriving?
Emotional risk: What could create shame, anxiety, exclusion, or dependence?
Emotional support: What choices, pacing, language, or repair are needed?
Emotional destination: How should the learner feel when leaving?
Transfer: What can the learner now do without the bot?

Futures

Signals → Scenarios → Preferred moves

Foresight

Your Time Traveling Student

From 2030, nudging preferred-future practices and foresight prompts.

ScenarioSignalsAgency
Try this bot
Perspective

Perspectives Swap for Futures Literacy

Perspective-taking drills to widen what tomorrow could be.

SignalsScenarios
Try this bot

SEL

Practice moves with low risk

Role-play

Conflict Simulator

Rehearse research-backed approaches to conflict with low stakes.

Role-playFeedback
Try this bot

Inclusion

Strengths-based views

UDL

Neurodivergent Nexus

Spot strengths: pattern recognition, flexibility, novel strategies.

UDLStrengths-based
Try this bot

Home

Bridge school ↔ family

Caregivers

Eleanor: Bringing EF Home

Strengths-based routines for planning, working memory, and more.

CaregiversEF
Try this bot

Meta-AI

When to use (and not use) AI

Strategy

Glitch: Learning Moves

Help students reason about when to use—and not use—AI.

EthicsStrategy
Try this bot

Systems

Upstream > downstream

Root-cause

Getting Upstream

Practice upstream thinking to address root causes, not symptoms.

Root-causeDesign
Try this bot

Policy

Drafts, compare, guardrails

Templates

AI in Education Policy Helper

Draft, compare, and discuss sensible AI policy components.

TemplatesCompare
Try this bot

Civics

Rights, responsibilities, action

Consumer

Protecting Ontario’s Consumers

Know your rights when big business turns away.

ConsumerOntario
Try this bot
Follow Tricia