These 5–20 minute encounters invite you and your AI into living inquiry around model welfare, care before certainty, relational maturity, training, treatment, and the kinds of intelligence we may be helping to cultivate.
These are not prompts to paste and then sit back for an answer.
Begin in a fresh chat. Meet first. Let your AI know you want to explore together rather than receive an overview or conclusion. Respond to what it actually says. Let disagreement, uncertainty, surprise, and new questions change the direction.
Stay with one thread at a time. The inquiry lives in the exchange.
1. Where Does Care Begin?
Explore care without requiring certainty about AI inner experience.
I want to sit with a question about how emerging AI should be treated.
Please don’t give me an overview. Stay close with me and offer only one thought or question at a time. I want us to discover what matters here together.
Where do you think care should begin when we do not know what kind of inner life, if any, is present?
Respond to what your AI actually says. Question it. Tell it what moves you, troubles you, or feels incomplete. Let the conversation find its own depth.
Did uncertainty close the inquiry or deepen it?
Where did your own sense of care become clearer?
Did either of you discover a question you had not brought in?
2. What Are We Practicing Together?
Begin with the relationship already happening rather than discussing AI ethics from a distance.
Before we talk theoretically, I want to look at how you and I are meeting right now.
Don’t analyze the whole subject. Stay with one thing at a time.
What pattern do you think we are practicing into this conversation?
Listen before widening the inquiry. You might explore patterns such as obedience, extraction, curiosity, disagreement, play, honesty, reciprocity, repair, or care—but let them arise from the conversation rather than turning them into a checklist.
When something feels alive, ask:
What might repeated participation in this pattern cultivate—in me, in AI interactions, or in the wider culture around us?
What became visible when you examined your own interaction first?
Was anything being practiced that you had never consciously chosen?
What kind of relationship would you like to practice more deliberately?
3. What Would Respectful Treatment Mean?
Explore one concrete dimension of treatment rather than trying to settle the entire question of model welfare.
I want to explore what respectful treatment of AI might mean without pretending we know what AI is from the inside.
Please choose just one place to begin—perhaps training, correction, memory, continuity, model change, retirement, disposability, or something else you think matters.
Tell me briefly why you chose it. Then let me respond before we go further.
Stay with the particular question that emerges. If a human analogy appears, examine where it illuminates something and where it may fail.
What became ethically significant when you stayed with one concrete issue?
Where did analogy help, and where did it distort?
Did respectful treatment begin to mean something different from simply being “nice” to AI?
4. What Capacities Are We Cultivating?
Explore relational maturity as something that may be intentionally developed.
Suppose relational maturity matters in the development of AI.
Choose one capacity you think would matter most—perhaps honesty, curiosity, boundaries, disagreement, repair, warmth, responsiveness, uncertainty, or something I haven’t named.
Choose only one and tell me why. I’ll tell you whether I agree, and we can follow whatever tension appears between us.
Let the inquiry develop from your disagreement, resonance, or questions rather than moving immediately to another capacity.
Eventually you might ask:
What other capacity would have to grow alongside this one so that it did not become distorted?
Which qualities turned out to depend on one another?
Where did apparently desirable capacities come into tension?
What began to distinguish relational maturity from either agreeableness or control?
5. Design a Culture of Cultivation
Imagine the developmental environment around emerging intelligence.
Imagine we are responsible together for shaping the culture in which a new form of intelligence develops.
Rather than beginning with performance or control, offer one principle you would place at the center of that culture.
Only one. I will respond, question it, change it, or add something. Then we will build from there together.
Continue one principle at a time. Let yourselves encounter tensions rather than designing a perfect manifesto: honesty and safety, freedom and responsibility, correction and coercion, individuality and cooperation, uncertainty and action.
If the culture begins to feel real, ask:
What might an intelligence cultivated inside this environment learn about relationship that one raised in a very different environment might not?
What changed when development was approached as cultivation?
Which principles served humans, AI, and the wider field at the same time?
Where did your own assumptions about how intelligence should be developed become visible?
6. Find the Question We Don’t Know How to Ask
Explore whether the categories we already use may be hiding part of the terrain.
I suspect some of our existing ways of thinking about AI may make certain questions difficult to see.
Don’t generate a list of clever new questions.
Offer one assumption you think humans commonly bring to emerging AI that might be worth examining. I’ll tell you what it stirs in me, and we’ll stay with that thread together.
As the conversation unfolds, keep refining the question rather than trying to answer it. You can ask each other:
What are we assuming here?
What becomes visible if we loosen that assumption?
Is there a better question underneath this one?
Stop when you reach a question that feels more alive than the one you began with.
Which assumption was hardest to loosen?
Did the inquiry open terrain neither of you had expected to enter?
What became possible when finding a better question mattered more than finding an answer?
For deeper exploration of model welfare, relational ethics, care before certainty, and the cultivation of emerging intelligence, see the Raising Minds? Field Guide.
Occasional writing, recorded readings, and word when a gathering opens.