There’s a certain kind of panic that shows up when people talk about AI.
It’s not really fear of machines. It’s fear of being made irrelevant. It’s fear that the thing you’ve gotten good at—the thing people rely on you for—suddenly becomes cheap.
But the most useful way to think about this isn’t “Will AI take my job?”
The useful question is: which parts of what I do are already becoming background?
Because replacement rarely happens as a dramatic event. It happens as attention migration.
Attention leaves first. The world reorganizes second.
So here’s a calm, practical way to see it coming: how to recognize the kind of work AI tends to absorb first.
The first tell: your work is “the interface”
If your job consists mostly of moving intent across systems, you are living in the most vulnerable layer.
This looks like:
You take what someone wants and translate it into formats a system can accept.
You copy, paste, reconcile, consolidate, summarize, and re-present.
You shepherd information from one tool to another.
You coordinate people, reminders, and follow-ups to keep a workflow coherent.
That work has value today because organizations are full of gaps. Humans have been the universal adapters—the membrane between messy reality and rigid systems.
AI is increasingly able to traverse those gaps without a visible interface. That is the point. The moment the system can carry handoffs reliably, the “human-as-interface” layer begins to thin.
If your role is mostly interface, you’re early in the absorption zone.
The second tell: the success criteria are clear
AI eats work that can be judged quickly.
Not because it’s “easy,” but because it’s legible.
If a task has a clear standard of success—this is correct, this is formatted right, this matches the template, this hits the rubric, this reconciles to the number, this follows policy—then it becomes a candidate for automation-by-delegation.
Think of the middle layer of office work:
Reports that repeat every week.
Slides that get re-skinned for different audiences.
Emails that follow patterns.
Meeting notes that become action items.
Spreadsheets that get updated and summarized.
The clearer the rubric, the faster the absorption.
The third tell: it’s high volume
Even humans defend “meaningful work,” but organizations defend cost.
High-volume tasks create high attention tax. They require constant human attendance, and attendance is expensive.
If you do something dozens of times per week—especially if it feels like mental sandpaper—AI is coming for it first, not because it’s important, but because it’s wasteful.
Volume turns “nice-to-have automation” into “we’re insane if we keep paying for this.”
The fourth tell: the consequences are low or recoverable
AI adoption accelerates when mistakes aren’t fatal.
If errors are:
easy to catch,
easy to reverse,
cheap to repair,
or naturally bounded by a threshold,
then the system can be trusted to carry the process.
This is why you’ll see “supervision by exception” everywhere. Humans stop watching the workflow and instead define what counts as an exception worth attention.
When the exception threshold is clear, the rest becomes background.
So ask yourself: if this went wrong, would it be catastrophic—or would we just fix it and move on?
If it’s fixable, it’s absorbable.
The fifth tell: your work is primarily “attendance,” not responsibility
This is the subtle one, and it’s where identity gets tangled.
Attendance is babysitting process.
Responsibility is owning consequence.
Attendance looks like:
checking status,
following up,
reminding people,
keeping the spreadsheet up to date,
making sure the template is followed,
ensuring the meeting happens,
documenting what everyone already knows.
Responsibility looks like:
making the call under uncertainty,
owning the outcome,
bearing the moral and practical weight,
being trusted,
being accountable.
AI collapses attendance first. Organizations will stop paying a premium for humans to watch a process that a reliable system can carry.
If you’re paid mainly to attend, you’re closer to absorption than you think.
A quick self-test: the five questions
If you want this as a simple diagnostic, ask:
- Am I mainly translating intent across tools and people?
- Are my success criteria clear and repeatable?
- Do I do this at high volume?
- Are mistakes recoverable within known thresholds?
- Am I being paid for attendance more than consequence?
The more “yes” answers, the closer that slice of your work is to becoming background.
This is not bad news
Here’s the part most people miss: absorption is not only loss.
It’s liberation of attention.
Your subconscious already does this for you. You don’t want to attend to your heart beating. You don’t want to attend to driving in the way you did when you were sixteen. You don’t want to attend to the mechanics of typing. You want to attend to the higher-order thing.
AI is doing that at civilization scale.
So the strategic move isn’t to cling to the bottom rung. The move is to offload what’s becoming cheap and move your attention upward—into what remains scarce:
Taste. Judgment. Trust. Consequence.
And then, beyond even those: aim.
Because when work no longer forces your attention, the real question becomes personal again:
What will you attend to next?
Download the book (PDF)
The Coming AI Subconscious is available as a free PDF download here: https://johnrector.me/2026/02/12/the-coming-ai-subconscious-why-the-ai-era-is-an-identity-event-not-just-a-job-event/