Co-Creating with AI

Transcript · Download MP3 · Download MP4

Invitation

AI can feel like a tireless assistant, an astonishing creative catalyst, an unreliable know-it-all, an economic threat, an ethical mess, a shortcut, a mirror, or a new participant in the human story.

Often it feels like several of those at once.

Co-creation is an attitude and a practice. It says: I am not creating in isolation. I am bringing my discernment, experience, values, reactions, and heartistry into relationship with something that can influence what emerges.

With another human, relationship includes bodies, needs, consent, history, accountability, and genuine mutual experience. An AI system is not a human co-creator in that sense. It does not love us, assume moral responsibility, possess our lived context, or know truth simply because its language sounds assured.

Yet interacting with it can still become a creative dance. It can offer possibilities at four in the morning, organize complexity, translate, compare, challenge, summarize, generate images, help us find language, and give us something concrete to react to when the blank page has become a wall.

The skill is not using AI as much as possible. The skill is developing a right relationship with it—one that increases human agency, emotional freedom, creativity, and responsibility rather than quietly replacing them.

This Guide is grounded in a January 2025 workshop. Products, capabilities, costs, and policies will keep changing. The durable work is learning how to meet powerful change without surrendering either our curiosity or our discernment.

Make Room for the Real Reaction

New technology can activate the primitive brain before we have enough experience to evaluate it.

We may fear becoming obsolete, losing work, being unable to learn, contributing to harm, watching human creativity become cheapened, or entering a future we did not consent to. We may also feel dazzled, relieved, greedy for speed, or eager to avoid tasks that have drained us for years.

Neither panic nor enchantment is a particularly stable place from which to choose.

Emotional freedom does not require deciding AI is good. It gives us room to feel the threat response, grief, excitement, and uncertainty without letting one reaction become the whole frame.

We can ask:

• What exactly am I afraid will happen?
• What am I grieving or protecting?
• Which opportunity feels genuinely useful?
• What boundary would let me experiment without betraying what matters?
• What consequences need more than optimism?

Tapping: From Threat Reflex to Curious Discernment

Cotton Candy, Green Beans, and Human Nourishment

Powerful new tools often arrive as cotton candy: bright, immediate, almost magical. We gorge on novelty. Images appear in seconds. A polished answer arrives before we have finished formulating the question.

Then the body asks for green beans—something with substance.

We still long for authentic warmth, trustworthy expertise, accountable relationship, a trembling human voice, shared laughter, and the feeling that another being is actually with us. AI may imitate some signals of human connection remarkably well. That makes our discernment about real relationship more important, not less.

The most useful role for AI is often not replacing the nourishing part. It is clearing some of the cognitive clutter around it:

• sorting a field of options;
• summarizing material we will evaluate;
• turning an emotional rant into the first draft of a clear request;
• producing alternatives that reveal what we actually prefer;
• handling tedious transformations so our human attention can return to relationship, judgment, and craft.

Speed is valuable when it returns life force to what matters. Speed is costly when it helps us bypass the very engagement that develops voice, wisdom, or connection.

The Intelligence Dance

If we type a vague request, receive the first output, and paste it into the world, we have used generation. We have not done much co-creating.

The dance begins when we remain present enough to influence what happens next.

1. Orient

Name what matters before asking for output:

• Who is this for?
• What should they understand, feel, or be able to do?
• What values and boundaries must remain visible?
• What would make the result useful rather than merely impressive?

2. Offer Context

AI responds to what reaches the conversation. Share relevant constraints, examples, source material, audience, desired tone, and what you have already rejected. Do not share confidential or identifying material merely because more context might improve the answer.

3. Receive Without Surrendering

Treat the response as material on the worktable.

Notice your reaction:

• That phrase has energy.
• This sounds generic.
• It missed the sacred part.
• I would never have thought of that.
• This claim needs evidence.
• I dislike all three options, and now I know why.

Reaction is not a failure of prompting. It is part of the intelligence you bring.

4. Influence

Tell the system what landed and what did not. Ask it to challenge assumptions, create contrasts, find missing perspectives, explain uncertainty, simplify, deepen, or preserve specific language.

Iteration is where a generic answer may begin responding to a living human intention.

5. Verify

Fluency is not evidence. AI can be warm, articulate, and wrong.

The workshop used the image of a beloved elder who was generous and helpful—and occasionally very convinced of facts that were not true. Holding that image can soften both extremes: we do not need to despise the helper, and we should still check the answer.

Verify consequential facts, quotations, calculations, sources, product details, medical or legal claims, and anything another person may rely on.

6. Integrate

The human decides what leaves the workspace. Rewrite. Combine. Add lived texture. Reject the output. Name AI assistance honestly where provenance matters.

The final question is not Did AI write this? It is Am I willing to stand for what I am putting into the world?

Tapping: Keeping My Voice in the Loop

Practical Magic: Give the Tedious Parts a Partner

The workshop offered ordinary examples because usefulness becomes easier to judge in real life.

AI can help compare appliances that must fit an exact space, locate a service provider, narrow entertainment choices according to a specific mood, organize research, draft a firm but professional letter, or explain a concept in another way when the first explanation did not land.

The magic is rarely that no human could perform the task. It is that searching, sorting, and drafting may consume energy we want for something else.

Try requests with lived constraints:

• “Give me three energetic, warm films available tonight. No horror, humiliation, or bleak endings.”
• “Turn these angry notes into a clear request. Keep the boundary firm; remove insults and invented motives.”
• “Compare these options using only the specifications I provide. Tell me what information is missing.”
• “Explain this concept three ways: visual, practical, and suitable for a curious teenager.”

Then check. Availability changes. Specifications are misread. A professional-sounding letter can still make a poor claim. The tool reduces effort; it does not inherit responsibility.

Learning What Once Felt Unlearnable

Many people once believed they would never use video calls, smartphones, online banking, or word processors. Humans have already learned an astonishing number of interfaces and conventions.

Fear of learning AI may contain several different fears:

• I cannot understand the technology.
• I will look foolish while learning.
• The tools will change as soon as I catch up.
• Learning it means endorsing it.
• If I become good at this, what happens to my old identity or profession?

We do not need mastery to begin. One prompt, one correction, and one low-risk task can teach more than hours of abstract dread.

Tapping: I’m Afraid I Can’t Learn This

Work, Worth, and the Fear of Replacement

AI is changing tasks and roles unevenly. Some work will become faster, some will be devalued, new work will appear, and some people will bear much more disruption than others.

Historical comparisons can help us remember that roles evolve. They should not be used to dismiss present fear. A person facing lost income does not need a lecture about typewriters. They may need practical support, policy change, new skills, grief, anger, and time.

Market demand is not a measure of human worth.

At the same time, refusing to look can leave us with less influence over the transition. We can explore:

• Which parts of my work require trust, accountability, relationship, embodied judgment, or lived context?
• Which repetitive tasks could become easier?
• What skill becomes more valuable when generation is abundant—curation, verification, taste, facilitation, care, strategy?
• Where do I need boundaries so speed does not become an expectation that depletes everyone?

The goal is not cheerful adaptation at any cost. It is more agency inside a change we did not individually choose.

Human Wisdom Is More Than Answers

People once reached toward parents, elders, teachers, librarians, friends, and practitioners for answers. AI may provide information faster than any one person can.

But relationship with a wise human was never only information retrieval.

A human may know our history, notice our face, bear responsibility for advice, challenge the question itself, sit with grief, tell a story from lived experience, or refuse to turn a sacred uncertainty into a listicle.

AI can supplement human relationship. It can also become an appealing way to avoid vulnerability, disagreement, waiting, or asking for help.

Watch the energetic pattern:

• Am I using this because it expands my capacity?
• Or because a system cannot be disappointed in me?
• Am I gathering perspective before talking with someone?
• Or replacing the conversation that actually needs to happen?

There is no universal correct answer. There is a relationship to notice.

Ethics Belongs Inside the Co-Creation

Ethics is not an appendix we add after enjoying the output.

Truth and Accountability

AI can invent facts, sources, quotations, and confident explanations. Verify what matters. Say when uncertainty remains.

Privacy and Consent

Do not share client stories, personal identifiers, confidential business material, health details, or another person’s private writing merely because the tool makes processing convenient. De-identification helps but does not solve every risk.

Authorship and Appropriation

Generated language is shaped by patterns learned from human work. Plagiarism checkers cannot certify ethical originality. Keep source material and provenance clear. Do not present generated synthesis as verbatim human speech.

Bias and Flattening

Systems can reproduce dominant assumptions and make culturally specific realities sound generic. Ask what perspective is missing and seek people with lived knowledge.

Environmental Cost

Training and operating large systems consumes energy, water, hardware, and data-center capacity. Efficiency changes over time; so does demand. Decide whether a use is valuable enough to justify its cost rather than assuming every generated convenience is free.

Power and Access

The benefits and harms of AI are not distributed evenly. Companies, workers, artists, communities, and countries have different power to consent, profit, refuse, or recover. Personal savvy matters, but it does not replace collective responsibility.

Across Languages and Cultures

AI can make rough translation, captions, summaries, and cross-language exploration radically more accessible. That can help wisdom travel beyond the language in which it was first spoken.

It can also miss idiom, humor, history, sacred meaning, power, and regional difference. A fluent translation can be culturally wrong.

Label machine-generated translation honestly. Invite human correction. Use technology to open the bridge while respecting the people who know how to cross it.

Practices for Living This Skill

Choose One Low-Risk Experiment

Pick a task where an error would be inconvenient, not dangerous. Let the purpose be learning the dance rather than proving AI is wonderful or terrible.

Ask for Options, Not an Oracle

Request three possibilities and the tradeoffs among them. Notice which option activates your own knowing.

Use the Reaction Loop

After every response, write:

• What feels alive?
• What feels false or generic?
• What is missing?
• What needs verification?
• What would I change if no one could see the original?

Protect a No-AI Zone

Choose activities where the wrestling, slowness, privacy, or human contact is part of the value. Not every friction needs removal.

Verify One Claim Deliberately

Follow a citation to the actual source. Check whether it says what the answer claimed. Build verification as a habit before the stakes are high.

Track the Energy Result

After using AI, notice: Do I have more agency and life force? Am I clearer? Or do I feel passive, scattered, ashamed, compulsive, or detached from my own voice?

Name the Provenance

When material becomes public or consequential, record what came from humans, source documents, AI synthesis, and your own review. Clear provenance protects trust.

Bringing It Into Your Life

Co-creating with AI is not a one-time decision for or against technology. It is an ongoing practice of relationship, boundaries, experimentation, and responsibility.

Some uses will feel like liberation from hours of cognitive clutter. Some will produce cotton candy that looks nourishing until we taste it. Some will cross ethical lines or ask us to surrender too much privacy, craft, or human contact.

We do not need a permanent verdict before we begin paying attention.

Keep your hands on the values. Keep your heart in the work. Let curiosity open possibilities, let verification test them, and let your body tell you when speed has started pulling you away from yourself.

The tool does not get to decide what thriving means.

Let AI earn its place in the co-creation—and remain willing to change the dance when it does not serve what matters.