The New Minimum

“I haven’t really looked into it.”

For most of my life, that sentence was explanation enough. Of course you hadn’t looked into it — looking into it cost an afternoon. Now a competent briefing is thirty seconds away, and I can feel the sentence changing in my mouth: less a fact about my time, more a confession about my choices. Nobody demanded this. No rule was announced. The cost of knowing collapsed, and what counts as reasonable moved with it.

It isn’t only knowing. You buy something without comparing twenty alternatives. You hear a claim without checking it. You enter a meeting without asking for the background first. You send the three-line reply instead of the polished one a machine could write in ten seconds. None of these used to be failures, and most of them still aren’t. But they can begin to look like failures once the cost of doing more approaches zero.

No one has to decide this. A capability becomes cheap. Enough people use it. Polished becomes ordinary; thorough becomes ordinary. Your unchanged work now sits beside work the machines helped polish, and the difference invites a story: they didn’t bother. They came unprepared. They could have checked. Nothing about your reply got worse. It simply fell below a standard that didn’t exist before.

This is how the new minimum moves: once doing more becomes easy, not doing more begins to require an explanation. When extra effort was expensive, it usually had to justify itself. “I didn’t have time” protected enormous parts of life from optimization. As that excuse dissolves, omission starts to look less like limitation and more like choice. And choice is much easier to judge than limitation: “I couldn’t” earns sympathy; “I didn’t” invites judgment.

But the machines do not decide what we owe each other. They expand what is possible. What is expected of us is set by people — by what we praise, what we excuse, what we hold against each other — and nothing about that is a law of nature. And “could have done more” is not the same as “more would have mattered”: every reply could be longer, every decision more researched, every document polished once more, and long before the machines run out of more, the improvements stop having a meaningful impact. What is possible is becoming bottomless. What we owe each other is not.

Some minimums should rise. I no longer guess at things I can check in seconds — that excuse has genuinely expired. The mistake is not rising standards. It is letting the existence of a capability decide, all by itself, which standards rise — treating “easy now” as if it meant “required now.”

So here is a small discipline for the next time AI makes something newly possible: don’t only ask, “Why wouldn’t I do this?” Ask, “What would justify making this the new minimum?” And when what you could have done starts to make you feel inadequate, hold the two sentences apart: Yes, I could have done more. No, that fact alone does not establish that I should have.

The frontier can move on its own. The minimum should not.