The product world agreed on the moat last week. It forgot to say whose.

Last week three product conferences ran in London inside four days. Mind the Product’s Leadership Forum on Monday, mtpcon at the Barbican on Tuesday, the Product-Led Festival on Thursday. Different organisers, different crowds, and one argument repeated from nearly every stage.

The argument: AI has made building cheap, so the thing worth paying for now is knowing what to build. Judgment. Taste. The decision, not the execution.

Christian Idiodi opened mtpcon by saying product management has “drifted from invention into administration.” John Moriarty from Fin put numbers to it, that something like 94% of his company’s pull requests are now written by AI, and that whole categories of product are ceasing to be worth building as a result. April Dunford made the case for positioning in a market where products ship overnight. At the Product-Led Festival, one talk was billed as “PM 2029: Not the CEO, but a Solopreneur.” (The quotes here come from Mind the Product’s own recap of the day.)

We think they are right. If you still measure a product team by how much it ships, you are measuring the one thing that just stopped being hard. The correction is overdue.

But listen to how the new moat gets described, and it is always one person. The product creator. The solopreneur PM. The individual with good instincts and a fleet of agents doing the building. Every framing put judgment inside a single head. Which is a fine story if you work alone, and most of the people in those rooms don’t.

Judgment is the hardest thing a team shares

Of everything a product team produces, judgment is the one thing that doesn’t pass cleanly between people. The code travels. The designs travel. The spec travels, more or less. The sense of why this and not that, the instinct that took thirty customer conversations and one bad launch to form, stays with whoever formed it. You can write the decision down. You can’t write down the thing underneath it.

We have written before about what happens when every person on a team builds their own AI setup: the team gets individually correct answers that contradict each other without ever colliding, and each person defends their own because the AI walked them through the reasoning. The conference consensus is the same problem viewed from the other end. If judgment is the moat, and judgment lives only inside individuals, then a team of fifteen has fifteen private moats that don’t add up to anything shared.

One speaker came close to naming it. Pendo’s Dave Killeen had the line everyone repeated, that the product manager has gone from Michelin chef to “truffle hunter, pattern-matching at scale.” The less-quoted part was sharper: the constraint now isn’t how productive you are, it’s coordination. When everyone can build, the question stops being whether you can build the thing and becomes whether the team agrees on which thing, and why, and whether the agents they have each set loose share the same understanding of the problem.

That reframes Idiodi’s line. The drift “from invention into administration” isn’t a failure of nerve. The status meetings, the re-explaining, the document nobody enjoys writing: that is the work a team does because judgment won’t spread on its own. It is the cost of keeping a shared picture alive by hand. Telling everyone to be more inventive doesn’t lower that cost. It raises it, because now there are more confident individual views to reconcile.

”Have more taste” doesn’t fix a coordination problem

And confident individual views are exactly what this AI moment produces. Spend an hour with an agent and you come out more sure of yourself, because the thing spent the hour agreeing with you. Run that across a team. Six people, six private AI conversations, six well-reasoned and slightly different readings of what the customer needs. Nobody is wrong, exactly. They have each been talked into their own version. The meeting doesn’t feel like a disagreement, which is the dangerous part, because everyone leaves assuming they’re aligned.

A fireside at mtpcon, with speakers from Figma and the MONY Group, made a point that fits here: AI didn’t break planning or quality or decision-making, it just exposed where teams were already weak. The cracks were always there. A team could carry mismatched mental models for years when shipping was slow, because slowness hid them. Six weeks of build time is six weeks in which someone notices the spec is wrong and fixes it before anything ships. Take the build down to an afternoon and there is no time for the gap to surface before the work is live.

So the useful question isn’t “does your team have good judgment?” Almost everyone says yes. It’s whether that judgment is something the whole team can see and use, or something locked in a few people’s heads.

Collective judgment as infrastructure

This is the part nobody put on a slide, and it is what we are building at Quack Stack.

Shared judgment isn’t a document you wrote once. It is a living picture of your market, your customers, your competitors, and the reasoning behind your decisions, kept current as new signals arrive, with the evidence still attached. Your PM asks in Slack what the biggest objection from your core segment is this month, and gets an answer grounded in the last few hundred customer signals. Your engineer asks the same thing from their IDE and gets the same answer, with the same evidence. The experiment the team is about to run traces back, through the opportunity it came from, to the customer evidence behind it. And the agents now doing a chunk of the work read from the same picture instead of each inventing their own.

A shared picture can still be wrong. The difference is that when it’s wrong, it’s wrong in one place that everyone can see and argue about, and everyone updates together. Twenty private versions that each feel right to the person holding them are wrong in twenty invisible ways at once.

The conferences were right that judgment is the moat. The harder truth, for anyone who isn’t a company of one, is that judgment has to be held in common. That is the problem worth solving, and it is the one the solo-operator story walks straight into the day that solo operator hires someone.

If that is the kind of team you’re running, we would love to show you how it fits. Book a demo.