You tell an agent what you need. It compares, selects, buys. You receive the item. Did you choose? You formulated a need and approved a result, which is no small thing. But the stage where you hesitate between two options, where you discover you'd actually prefer something else, has disappeared. And that stage did something we had never needed to name.
This morning we reported on an agent designed to operate right through the purchase journey. Let's take the time to consider the question it raises.
What comparing produced
Comparison is presented as a chore, and it often is one. But it did three things, none of which is being replaced.
It sharpened the need. You think you know what you want before you look. In practice, it's by seeing the options that you discover what matters: that feature you didn't know existed, that trade-off you hadn't considered. The need forms in the comparison; it doesn't entirely precede it.
It built a skill. Someone who has compared ten cameras knows something about cameras. That knowledge serves the next choice, and it makes you less dependent.
It committed you. A choice you have weighed, you own. A choice you have been handed, you submit to. Satisfaction with a purchase comes partly from having decided, not only from the object.
The serious objection
It has to be stated, because it is strong.
Comparison as it exists today is nothing noble. It takes place in interfaces designed to steer, with paid rankings, dubious reviews and an abundance calculated to exhaust you. We noted this in relation to web search: on heavily optimised subjects, the raw material available is of poor quality.
Under those conditions, delegating to an agent capable of reading a hundred spec sheets without being distracted may produce a better choice than the one you would have made. That is not a flattering hypothesis for us, and it is probably right in many cases.
A comparison tool that works for you and one paid by those who sell do not do the same job, even if they show the same interface. We noted this morning: an agent offered by a company whose revenue comes from advertising operates in a particular configuration. That is not an accusation, it is a question to ask systematically. An agent that compares on your behalf occupies exactly the position once held by comparison sites, buying guides and rankings — and that position has always been contested by those with an interest in the outcome.
What distinguishes the two situations
Not everything is equivalent, and the useful criterion is the recurrence of the object in your life.
For a consumable, a spare part, an item where only conformity matters: delegating is excellent. Comparison teaches you nothing, preference does not exist, and the time saved is clear.
For something you will use every day for years — an instrument, a work tool, an object you will come to master — comparison is part of making it yours. You are not just choosing an object, you are beginning to learn that domain.
This is the distinction we proposed in our article on what deserves to be automated: the question is not the time saved, it is whether the task teaches you something.
The long-term risk
There is a consequence that goes beyond the individual purchase.
A preference you no longer exercise atrophies. If a system always chooses for you according to what it has understood about you, it offers you a frozen version of who you were when it learned. You stop discovering that your tastes have changed, because nothing puts you in a position to notice any more.
It is the same mechanism we described in our article on consensus: the problem is not that something is imposed on you, it is that the effort required to encounter something else disappears.
What to take away
The answer to the question in the title is nuanced: yes, formulating a need and approving a result is a choice. But it is a poorer choice than the one that goes through comparison, because it cannot change your mind.
A choice that cannot surprise you is not quite a choice: it is an order. The difference is slight on a packet of coffee, and considerable on what will stay with you for a long time.
The right reflex, then, is to ask yourself before delegating: could I discover something by looking myself? If the answer is no, delegate without hesitation. If it is yes, the time you lose is not wasted.