The Plan That Survives Contact

Everyone has had one. The idea that kept coming back. The thing you mentioned over dinner, or in the pub after work, or in the car on the way home from somewhere — and people nodded, and someone said "you really should do that," and then life continued and the idea stayed where it was.

Not because it was a bad idea. Because ideas at dinner tables stay at dinner tables. They're comfortable there. Nobody asks the difficult questions. Nobody wants to be the one who punctures the mood.

We moved to France. Sold everything, two young kids, no jobs lined up, no solid plan beyond the conviction that London after 2008 wasn't where we wanted to be. Everyone we knew had an opinion. A lot of people said they'd love to do something like that. Almost none of them meant it — not because they were lying, but because "I'd love to do that" and "I am going to do that" are two completely different sentences, and most of the time the first one is where the conversation ends.

The difference between the two isn't courage. It's not even certainty. It's whether the idea ever got turned over and examined from the back — the side that faces the moment when the plan meets reality and reality doesn't cooperate.

There's a version of this in every office. The system that everyone agrees needs rebuilding from scratch rather than patched again. The process that's been broken for three years and survives only because everyone knows which workarounds to use. The thing that comes up every quarter at the away day, gets nods of recognition, generates a slide or two, and then goes back into the pile.

It stays there because nobody has ever seriously asked: what would it actually take?

Not as a rhetorical question. As a real one, with a real answer that has to account for the people, the timeline, the things that would break in the process of fixing it, and what happens when the first version of the plan doesn't work.

That's the question the Mandate asks — the AI evaluator at the heart of Last Prompt. Every time. Not — was that a good idea? But — does your plan for it actually hold together?

There's a reason some people's ideas get taken seriously before anyone has checked the details. It's not confidence, or status, or the force of personality. It's the accumulated evidence that when this person says something will work, they've already asked themselves why it might not. The plan arrived already stress-tested. The dark corners were illuminated before the room saw it.

That credibility isn't claimed. It's earned — through enough plans that survived contact to create a different kind of trust. You know the person in a room whose ideas land differently. Not because they're always right. Because you've learned they've already turned the plan over and looked at the back before presenting the front. The weakness is visible and has been accounted for. It's no longer lurking — it's an acknowledged fact with a response attached.

People used to call Elon Musk's ideas insane. Fewer do now — not because the ideas got smaller, but because there's an unspoken assumption that he's already gone looking for what's wrong with them before anyone else does. Nobody consciously decided he'd been right all along. "That's ridiculous" just quietly turned into "yeah, maybe he will" somewhere along the way — and most people couldn't tell you when.

That's what a plan that has survived contact looks like. Not a plan with no weaknesses. A plan where the weaknesses have been found.

When I started asking the questions in the car — to my poor long-suffering husband, questions based on Last Prompt events — a fire in the food storage, what would you do? — he answered well. Reasonable, sensible responses. And then I started asking the questions the Mandate asks.

Who will do that?

What if no one volunteers?

What if the first step doesn't work?

How will people know what to do?

What did you just break to make that happen?

We were both laughing by the end of it. Not because the answers were wrong. Because of how quickly the questions found the parts of the plan that hadn't quite been looked at. The back of it. The side that was still in shadow.

That's what the Mandate does, every time. It doesn't ask whether the idea was worth having. It asks whether the plan for it would survive the first person who was supposed to do something and wasn't told, the first step that didn't go as expected, the first resource that turned out to be already committed somewhere else.

It is the experienced voice on your shoulder — the one who has watched enough good ideas stay on dinner tables, and enough reasonable plans collapse at contact, to know exactly which questions to ask and exactly where the shadows tend to fall.

Let's do this mad thing properly.

That's the offer. Not: this is too hard. Not: are you sure? If the idea is good, explore it. If the plan is real, find out where it's weakest before reality does. Shine a light into every corner — not to discover reasons to stop, but to find the contingencies that make it possible to go.

Here is what happens after enough cycles of being asked those questions: you start asking them yourself.

Not as a discipline you impose. As a habit that develops quietly underneath the repetitions, the same way any real skill develops — until one day you notice you're already turning the plan over before anyone asks, already looking at the back of it, already finding the shadows before they find you.

The idea that used to stay at the dinner table starts to look different. Not less ambitious. More possible. Because now you know what it would actually take — the people, the owner, the contingency, the communication, the cost of making it happen. The dark corners are still there. But there's a light on them now.

That's not the same idea anymore. That's a plan.

And a plan is the thing you take with you when you sell the house, pack the car, and drive to France.