"To infinity and beyond." Buzz Lightyear says it with total conviction, arm raised, as though he is reading a coordinate off a map. It is the most quietly absurd line in children's cinema. Infinity is not a place. You cannot arrive there. And "beyond" — beyond infinity — is not even a coherent direction. The line works precisely because it sounds like a destination while being nothing of the kind. It is a statement of intent wearing the costume of a navigational instruction.
Most dates are the same.
We say a date the way Buzz says infinity — as if we are naming a fixed point the work will arrive at. Delivery: 14 September. It has the grammar of a measurement. It looks like a fact about the future. But a date is almost never a fact. It is a promise dressed as a forecast, and the difference between those two things is where most of the trouble lives.
A forecast is a probabilistic claim. Given what we know now, the work will most likely finish somewhere in a range, with some confidence, and that estimate will improve as we learn more. That is an honest thing to say about an uncertain future. A promise is different. A promise has no error bars. It does not say "most likely." It says "this will happen, then." And somewhere between the planner's spreadsheet and the steering committee's slide, the forecast quietly becomes a promise. The range collapses to a single number. The confidence interval is deleted. The honest "we think, with maybe sixty percent confidence, sometime in the third quarter" arrives in the boardroom as a date in bold, with a flag planted on it. The uncertainty did not survive contact with the slide.
And here is the part that ought to bother us more than it does. Once the date is fixed, it stops describing the work and starts changing it.
Parkinson's law — work expands to fill the time available. Student syndrome — nobody starts in earnest until the deadline is close enough to frighten them. Give a team eight weeks for a thing that needs four, and it will take eight. Give them four for a thing that needs eight, and the date does not move — something else does. This is the quiet violence of the iron triangle: scope, time, cost. Pin the date and one of the others must give way. Usually it is quality, because quality is the constraint with no one in the room to defend it. The date holds. The work erodes underneath it, invisibly, to keep the promise intact.
So the date does not merely predict the outcome. It bends the outcome toward itself. The number we picked to describe the work becomes the thing the work is organised around — and those are not the same activity. Hitting the date and doing the work well drift apart, and on a long enough programme they can become outright enemies.
What makes this almost comic is that the date was wrong the moment it was written. We are systematically, reliably optimistic about how long things take — Kahneman and Tversky called it the planning fallacy, and it is one of the more robust findings in the study of human judgement. Hofstadter put it more sharply: it always takes longer than you expect, even when you take into account Hofstadter's law. The estimate is biased low before anyone has done anything. But because we converted it into a promise, being wrong is no longer an ordinary forecast update. It is a failure. It is someone breaking their word. We have taken the unremarkable fact that the future is uncertain and rewritten it as a story about broken trust.
Scale this up to large programmes and the pattern stops looking like an accident. Big infrastructure and IT programmes run over time and budget so consistently that overrun is closer to the rule than the exception — Bent Flyvbjerg has spent a career documenting it. His more uncomfortable observation is that the optimism is not always innocent. Sometimes the date is low on purpose. An accurate date — the one with the real range attached — would not get the thing approved. So a better-looking number is offered, because the number is what is actually being bought. The date sells the project. Then the project spends the rest of its life trying to become true to a figure that was never about the work in the first place.
That is the uncomfortable truth underneath all of this. We pretend dates are predictions because predictions are neutral — nobody's fault if the weather turns. But we use them as promises, because a promise is what closes the deal, wins the mandate, releases the budget. The forecast is the costume. The promise is the body underneath.
None of which means we should stop putting dates on things. Coordination needs a shared sense of when. The real question is what we are willing to admit a date is. To say "most likely October, possibly December, and here is what would move it either way" is honest, useful, and almost impossible to put in bold on a slide. To say "14 September" is clean, confident, sellable — and a small lie we have all quietly agreed to tell.
Buzz, at least, had the decency to be aiming at something that obviously did not exist. We aim at the fourteenth of September and call it realism.
So the question is not whether a date is achievable. It is what we are doing when we name one — describing the future, or trying to bend it to a promise we already needed to make.