The death of the checkbox
The tasks of the future won’t just be a list with checkboxes. Most likely, there won’t be checkboxes at all.
We’re used to a simple model: there’s a task, you go off and do it “on the side”, then you come back and tick the box. The tick is the proof. It says: “done.” But it says nothing about what exactly was done, why, or what came of it.
In a world where agents do a large share of the work, the tick loses its meaning. The value moves somewhere else.
What “doing a task” means now
Doing a task in the future means, first of all, explaining three things: what you’re going to do, what you’re doing, and what result you got.
Not pressing the right buttons. Not sitting out the right number of hours. But stating an intention, walking the one who carries it out — a person or an agent — through it, and honestly recording what came of it.
It sounds minor, but it shifts the whole logic of work. A task stops being an item on a list. It becomes a small story with a beginning, a middle and an end.
Hands-on work is disappearing
Fewer and fewer tasks will need “hands-on” work. By that I mean work inside tools — Figma, Amplitude and the rest. All of them will run through prompts, partly or completely.
A designer won’t drag rectangles around; they’ll explain what the user should feel. An analyst won’t assemble a chart; they’ll frame the question the chart has to answer.
And the main result of a person doing a task will be explaining to an AI or a robot what needs to be done.
Every task is a conversation
That’s why every task is communication: with people or with specialised agents.
It’s important to understand: there won’t be tasks where you can do something “on the side” and just come back to tick a box. We are, of course, talking about the future — but it isn’t that far away.
Think how often your manager or colleagues ask you: why did you do it that way? What did you get as a result? Now an agent will be asking. And it will ask not afterwards, but before you start doing anything — and right in the middle of the work.
This isn’t control for the sake of control. An agent can’t carry out what you haven’t explained. And an explanation that exists before the work starts is already half the result.
Communication is project management
If every task is a conversation, then a project is a conversation too. Teams think, argue, invent — and AI quietly weaves structure, synchronisation and measurable progress into that flow. People stay themselves, and the routine disappears.
| Before | Now |
|---|---|
| Pick a methodology and force ourselves to stick to the process | Talk naturally, and AI builds the framework that’s needed |
| Being organised is measured by discipline | Being organised is measured by creative contribution; AI provides the discipline |
| Follow-ups, minutes and canvases are made by hand, often late | Documents appear instantly, while the conversation is still going |
| Discipline is the main lever of effectiveness | Individuality without penalty: different paces, languages, time zones; AI evens out the chaos |
| Lost focus, “it drowned in the chat” | AI sorts the conversation into tasks, problems, decisions, questions and insights |
| Useful work stays invisible | Every step is visible: what’s done, where the blocker is, where value is created |
What it means already today
Even now, while agents aren’t everywhere yet, it matters more to understand what you want to do, for what purpose and what result you want, than to draw, count or code.
A person who can explain a task clearly becomes more productive than one who can execute it fast. Because execution is more and more often taken on by someone else. Explanation isn’t.
The checkbox isn’t dying because tasks are disappearing. It’s dying because “done” stops being an answer. The answer becomes “here’s what I wanted, here’s what was done, here’s what came of it.”