Artificial intelligence is one of those topics where the obvious answer is right about sixty per cent of the time, which is exactly often enough to be dangerous.
The objection worth taking seriously
The interesting constraint is almost never the one in the brief. Where a design is obvious the prose is short, so the length of an explanation is a reasonable proxy for where to look next. The version of this that works fits on an index card. The version that fails needs an onboarding session.
Scope is the variable everyone adjusts last and should adjust first. The decision is usually cheap and reversible; the execution is where the cost lives, and that is where the argument should have happened. The counter-argument deserves a hearing, and it is stronger than its usual proponents make it sound.
A shared definition of "done" removes more friction than any tool. The things that are easy to count are rarely the things that matter, and once a number reaches a dashboard it starts shaping behaviour whether or not it deserves to.
The incentive problem
Most of the difficulty lives at the boundaries, not in the middle. The stated constraint is usually a proxy for a real one nobody wants to say aloud, and optimising the proxy is wasted effort. A useful test: if this disappeared tomorrow, how long before anyone noticed?
Feedback loops shorter than the planning cycle change everything. Teams that pick both end up with neither, and usually discover this at the point where reversing would have mattered. The clearest signal was that people stopped asking where things were.
What the data actually shows
The compounding effects matter far more than the individual wins. Being right sixty per cent of the time builds exactly the kind of confidence that makes the other forty per cent expensive.
Consider the failure mode rather than the success case. Success has many causes and teaches very little; failure tends to have one, and it is usually obvious in hindsight. The evidence here is thinner than anyone quoting it tends to admit.
What would change our mind
Nobody gets credit for the work that did not need doing. Artificial intelligence rewards clarity here more than almost anywhere else, because the wrong target produces work that looks productive and moves nothing. In practice the answer showed up in the calendar before it showed up in the dashboard.
The first thing to establish is what you are actually optimising for. If you learn on Friday what you assumed on Monday, the assumption never has time to become an architecture. We ran both approaches in parallel for six weeks. The difference was smaller than the cost of the debate about it.
The default answer is right often enough to be dangerous. A team that changes approach every quarter pays a coordination tax that routinely exceeds whatever the change was meant to fix. Ask what would have to be true for the opposite approach to be correct, and see whether anyone can answer.
You can have it fast, or you can have it reversible. Pick before you start, not after.
— Overheard in a retrospective
A more modest claim
Speed and reversibility are the trade-off worth naming out loud. Handoffs between people who each hold a coherent local picture and no shared one produce most of the pain later attributed to tooling. Try writing the constraint on one line before opening a vendor comparison; the line is usually harder than the comparison.
The second-order effects arrive about a quarter after the first-order ones. The first quarter shows the intended effect; the second shows what the intended effect displaced. The caveat is that all of this assumes the underlying goal is settled, which is frequently the actual problem.
The received wisdom
It helps to separate the decision from the execution. Cutting scope early is cheap and slightly embarrassing; cutting it late is expensive and deeply embarrassing.
There is a version of artificial intelligence that is mostly ritual. Choosing infrastructure before agreeing what it is for is how organisations end up maintaining a system nobody wanted.
A different reading
Consistency is worth more than any individual improvement to artificial intelligence. A small improvement applied consistently beats a dramatic one applied once, which is unsatisfying advice precisely because it is correct. It is worth saying that we have not run this long enough to be confident.
Documentation is a symptom: you write it where the design is unclear. They are decisions made quickly, defended slowly, and built upon for six months before anyone recalculates.
Measurement is usually where this falls apart. When responsibility is spread across a group, the work that falls between the named parts is the work that does not happen. This is easier to write than to hold to when a deadline appears.
A few things worth checking before you commit:
- Decide in advance what would make you stop
- Keep the feedback loop shorter than the planning cycle
- Name one person accountable — not a group
- Prefer the reversible option when the evidence is thin
How we got here
The expensive mistakes here are rarely the technical ones. It is comfortable, it is legible to management, and it is close to worthless once you measure what it actually changes. Reasonable people land elsewhere on this, usually because their constraints differ more than the vocabulary suggests.
What looks like a process problem is frequently an ownership problem. Most disagreements that present as strategic turn out, on inspection, to be two people using one word for two things.
The tooling question is downstream of the constraint question. Subtraction is structurally underrated: the meeting that stopped happening leaves no artefact to point at in a review. When we mapped it out, four of the seven steps existed only to compensate for the second one.
Where this leaves us
A shared definition of "done" removes more friction than any tool. Teams that pick both end up with neither, and usually discover this at the point where reversing would have mattered. One team we spoke to cut their review stage entirely and found throughput unchanged, which told them something the metrics had not.
Scope is the variable everyone adjusts last and should adjust first. Cutting scope early is cheap and slightly embarrassing; cutting it late is expensive and deeply embarrassing. Set a date at which you will stop, and write down in advance what would make you stop earlier.
The objection worth taking seriously
The expensive mistakes here are rarely the technical ones. A small improvement applied consistently beats a dramatic one applied once, which is unsatisfying advice precisely because it is correct. Set a date at which you will stop, and write down in advance what would make you stop earlier.
Consistency is worth more than any individual improvement to artificial intelligence. It is comfortable, it is legible to management, and it is close to worthless once you measure what it actually changes.
Consider the failure mode rather than the success case. The decision is usually cheap and reversible; the execution is where the cost lives, and that is where the argument should have happened.
The incentive problem
The first thing to establish is what you are actually optimising for. When responsibility is spread across a group, the work that falls between the named parts is the work that does not happen. There are organisations where the opposite is true, and they are not obviously worse off.
Documentation is a symptom: you write it where the design is unclear. Handoffs between people who each hold a coherent local picture and no shared one produce most of the pain later attributed to tooling. A useful test: if this disappeared tomorrow, how long before anyone noticed?
The interesting constraint is almost never the one in the brief. The first quarter shows the intended effect; the second shows what the intended effect displaced. That said, none of this generalises cleanly across team sizes.
What the data actually shows
The compounding effects matter far more than the individual wins. If you learn on Friday what you assumed on Monday, the assumption never has time to become an architecture. Try writing the constraint on one line before opening a vendor comparison; the line is usually harder than the comparison.
It helps to separate the decision from the execution. They are decisions made quickly, defended slowly, and built upon for six months before anyone recalculates. The caveat is that all of this assumes the underlying goal is settled, which is frequently the actual problem.
None of this generalises perfectly. Take the parts that map onto your constraints and discard the rest — that is what the framing is for.
