The supermarket tomato is a triumph of engineering. It ripens evenly, ships without bruising, holds its colour under fluorescent light for days, and arrives at the shelf uniform in size and shade. By almost every metric a grower or distributor would name, it is a better tomato than the irregular, soft-shouldered fruit it replaced. It is also, as nearly everyone who has eaten one knows, a worse tomato.
This is not nostalgia. It is documented. Decades of breeding selected for the traits that could be counted, yield, firmness, shelf life, uniform colour. One of the things growers selected for was a mutation that made the unripe fruit ripen evenly to a pleasant pale green, easier to harvest and grade. The mutation switched off a gene. That gene, it turned out, was part of a complex that governed the production of sugars and the compounds that carry flavour. In optimising for the appearance of even ripening, breeders quietly disabled the machinery that made the tomato taste of anything.
No one decided to remove the flavour. That is the point. The flavour was never on the spreadsheet. It was not a target, so it was not protected, and it left through a door no one was watching.
The measured and the unmeasured
Efficiency is not a value. It is a ratio, output over input, and it can only operate on the quantities you feed it. Tell a system to maximise firmness and shelf life and it will do exactly that, with the obedience of a machine. It will not pause to ask whether firmness and flavour are carried by the same biology. It cannot ask. It knows only the variables it was given.
This is the failure that hides inside every optimisation: the unmeasured variable is not weighed against the measured one. It is simply absent from the calculation. It does not lose the argument. There is no argument. And because the result still looks like a tomato, red, round, sliceable, the loss registers only at the moment of eating, far downstream from the decision that caused it.
The lesson is not that breeding is bad, or that we should refuse to improve things. It is more precise than that. Optimisation is reliable when the thing you can measure is the thing you actually want. It becomes a quiet act of destruction when the measurable variable is merely a proxy, and the real value rides along, unnamed, in the part of the system no one thought to protect.
Processes that cannot be hurried
Some things resist measurement because their essential quality is a function of time, and time cannot be compressed without changing the result.
Curing is one of them. Cold-process soap is chemically complete within a couple of days, the oils and lye have reacted, the saponification has run its course. By the logic of efficiency, the bar is finished. It could be cut, wrapped, and shipped. It would be soap. It would also be soft, harsh, and quick to dissolve.
What happens over the following weeks is not chemistry in the dramatic sense. It is water leaving. The bar loses moisture slowly, hardens, and the lather changes character, becomes finer, milder, longer-lasting. None of this can be rushed with heat or a fan without coarsening the result. The wait is not inefficiency to be engineered away. The wait is the process. There is no version of the bar that is both finished today and good.
The same structure appears wherever timing is non-negotiable. Wine and cheese do not become themselves faster under pressure. A stock reduces at the speed it reduces. Wood seasons over years, and kiln-drying, which is faster, produces timber that behaves differently in the hand and over time. In each case the duration is not overhead attached to the real work. It is the work.
When optimisation defeats its purpose
The error is not efficiency itself. Efficiency has given us more food, less waste, more access, and it is foolish to pretend otherwise. The error is reflexive efficiency, optimisation applied without first asking what the thing is for, and whether the value being sought can survive the variable being maximised.
A shipping company should optimise for shipping. A tomato is not a shipping problem. When the two are confused, you get a fruit that travels beautifully and tastes of nothing, and the confusion is invisible until someone bites in.
So the question worth holding is not whether to make things faster. It is narrower and more useful: what, exactly, am I measuring, and what is riding along with it that I have not named? The flavour of the tomato was real. It was simply unaccounted for. Most of what matters is.
Some things should be efficient. Others should be left to take the time they take, because the time is not the cost. It is the substance.