A convenience always looks free at the point of use. That is more or less the definition of one. Somebody has arranged for the effort to happen somewhere else, out of sight, and the whole pleasure of the thing lies in not having to think about where.
Nothing instant is actually instant
The result that appears in a fraction of a second was produced by machines that ran, in buildings that were cooled, drawing power from somewhere. None of that is sinister. It is simply the arrangement, and it is worth looking at squarely rather than flinching from it, because an arrangement you can see is an arrangement you can improve.
What makes this an engineering question rather than a mood is that the costs are extremely uneven. Two tools that feel identical can differ by a wide margin in what they consume, depending on how much work is done each time somebody asks, how much is remembered from the last time, and whether the heavy part runs once or on every single request. Those are design decisions, and they are made early, usually by people who are not thinking about them in those terms at all.
Visible costs change behaviour
The most effective thing I have seen is also the least dramatic: put the number where the people making the decisions can see it. Not as a judgement, just as a figure beside the others. Teams that can see what a feature costs to run tend to make it cheaper without being asked, in the same way that a household with a visible meter tends to notice the thing left switched on in the spare room.
There is a pleasing side effect. Work done to reduce what something consumes nearly always makes it quicker as well, because both come from the same source, which is doing less unnecessary work. Careful engineers have known this for a long time. The habits are old ones: cache the answer you already worked out, do the expensive thing once instead of a thousand times, and be honest about which requests genuinely need the heaviest possible treatment.
A plainer question at the end of the month
None of this is an argument against convenience. Convenience is one of the genuine goods of modern engineering, and a great deal of it has given people back hours they would never otherwise have had. The question is narrower and more cheerful than that. It is simply whether a particular convenience is worth what it takes to produce, and whether the people choosing it can see enough to answer honestly.
October is a reasonable moment for that kind of accounting. The year has shown its shape, the quarter is closing, and the habit of asking what a thing costs to run, rather than only what it costs to acquire, is a good one to carry into a new month.
Nicole Junkermann keeps the month's shorter jottings in the notebook and what is coming up on the calendar. The related note on climate technology worth backing takes the longer view, and open standards and quiet progress went up on the twenty-eighth.