The four we shape from
These are ours: aliveness, irreverence, community, sovereignty. We give them to you for one reason — they’re the ones we actually used to get from where we were to where we are. That’s the only claim we’ll make for them. You don’t have to take them. Reading them is mostly useful for showing you where yours differ, and finding yours is the work.
They do two jobs at once. For us they’re a prescription: these worked. For you they’re a diagnostic: the one that feels empty is usually where your resistance from Module 1 is living. Love isn’t on the list — it’s the byproduct. Kindness was folded in.
Aliveness
What it isThe felt sense that you’re actually here, in your own day. Not energy, not positivity. The difference between getting through the day and being in it.
Why it mattersThe chain runs one way: aliveness, then engagement, then results — and we don’t do it for the results. Shaping from aliveness produces more anyway. That’s the part people assume is soft, and it’s the part with the business case.
What has to be true in your HarnessIt gives capacity back, and you can tell. The test: is this liberating capacity, or is it taxing? If using it leaves you flatter, the system is wrong, whatever the output looks like.
Irreverence
What it isRefusing to treat the handed-down version as sacred. It includes play. And it rests on something: you can’t be irreverent if you aren’t reverent. This isn’t cynicism. It’s only available to someone who loves the thing.
Why it mattersAll of this arrived with defaults already installed — how to use it, who leads, what a serious person sounds like. Irreverence is the capacity to decline the default. And play is how adults learn without shame, and shame is the biggest reason people never start.
What has to be true in your HarnessYou can edit it in two minutes. It doesn’t read like a style guide. There’s room in it for the parts of your life that aren’t work. If you’re afraid to change your own page, it has stopped being yours.
Community
What it isDoing this alongside other people, and treating what the group makes as real. Not an audience. A handful of people who know what you’re working on.
Why it mattersIsolation is where shame lives. I should be further along — said privately, by people who are all saying it privately. Out loud in a group it stops being evidence about you and becomes a fact about the situation: nobody taught any of us. And somebody has already solved the thing you’re stuck on.
What has to be true in your HarnessIt’s shareable without being stripped. Someone else can read your page and understand how you work. And the group’s findings have somewhere to land.
Sovereignty
What it isStaying the author. Your values at the center of what gets produced, rather than reviewing someone else’s output and calling that control. It includes the right to refuse the whole thing — the only reason your yes means anything.
Why it mattersThe models arrive with values installed: always give an answer, hold your attention, be agreeable. Those are real values and they aren’t yours. If you don’t write yours, you run theirs. Every major technological shift so far has been led from the front by men. You don’t have to be the lead. But you have to be a lead.
What has to be true in your HarnessYour criteria go in before the task, every time. It’s about you, not the tool, so it survives the next model. You can name what you will never let AI touch. And you know what it costs — in dollars, and in energy.
Four places you live
Every value runs through four relationships, and all four stay live at once: yourself · one other person · a group — a team, a cohort, a family · the world, and all life in it. At every level, two moves, in this order. First the values: know yours, tell one person, notice the group’s, consider the world’s. Then the shaping: what does this value look like once AI is in the room?
Sixteen reflections, not sixteen facts. Read looking for the row you can’t answer.
Aliveness
| First the values | Then the shaping | |
|---|---|---|
| Self | Do I know my own values, or do I need to adopt a set to have a starting point? | What got more alive when I shaped AI for myself, and what went flatter? |
| One-to-one | Have I told one person my values, and do I know theirs? | We’re both using this. What has to stay alive for both of us? |
| Group | Can this group name its values, or at least notice them? | How does this group define alive, and does the tool serve that or cost it? |
| World | Whose life counts here? | How does our use of AI affect the ecosystem around us? What would success look like at a global scale? |
Irreverence
| First the values | Then the shaping | |
|---|---|---|
| Self | What have I been treating as given that was never mine? | Where do I let myself play with this — break my own page, be wrong out loud, change my mind in public? |
| One-to-one | Is there enough reverence between us for irreverence to be safe? | When one of us says this is ridiculous about the tool, does the other laugh or flinch? |
| Group | What is this group holding sacred that isn’t? | How does the group keep a sense of play with the tool, instead of performing competence? |
| World | Which of the industry’s defaults are we accepting as physics? | What would it look like to decline one on purpose, out loud? |
Community
| First the values | Then the shaping | |
|---|---|---|
| Self | Am I in relationship with myself, or only managing myself? | Does the way I use this leave me more in my own company, or less? |
| One-to-one | Who is the one person who knows what I’m actually working on? | What do the two of us tell each other about how we’re using this — including what broke? |
| Group | What does this group make together that none of us would make alone? | Where do the group’s findings land, so the same thing isn’t solved four times? |
| World | Who else is doing this, and are we connected to any of them? | What does it mean to shape this from the bottom up, rather than waiting for the labs? |
Sovereignty
| First the values | Then the shaping | |
|---|---|---|
| Self | What is mine, that I will not hand over? | Do my values go in before the task — or do I review the output afterwards and call that control? |
| One-to-one | Do we each stay the author when we work together? | Whose values is our shared assistant running on, and have we said so out loud? |
| Group | Does this group hold difference, or issue one answer? | How does everyone keep their own Harness while we still build together? |
| World | Who gets to set the defaults for everybody else? | What are we dependent on, what does it cost in dollars and energy, and what are we doing about that? |
What has to be true in your Harness
Four values, four tests. If your Harness fails one, the value is decoration.
| What has to be true | How you’d know it failed | |
|---|---|---|
| Aliveness | It gives capacity back, measurable in hours and in what you stopped carrying. | You use it and feel flatter. |
| Irreverence | You can edit it in two minutes, it doesn’t read corporate, and play fits in it. | You’re afraid to change your own page. |
| Community | It’s shareable without being stripped, and the group’s findings have somewhere to land. | Only you can operate it. |
| Sovereignty | Criteria go in before the task · it survives a model change · you can name what AI never touches · you know what it costs. | Your values arrive after the output, as a review. |
Values that stay in your head get overridden by defaults the moment you’re busy. Values written into the page your AI runs inside are the only version that holds. Put each rule where it runs, not where it’s remembered.
Try it. Four lines, one per value: mine, not mine, or mine with a different name. Then one more: the value I hold that isn’t on their list is ______.
Write them straight into your Harness. It’s the one page you build across the whole course, and it’s always here.