Through Time

Work: tasks, careers and small organizations

How to break your job into tasks, decide what to learn next and count whether AI paid for itself in a small team.

A course for employees, people starting out and owners of small organizations. You will take your job apart into tasks, draw up a plan for what to learn, agree on rules in a team and run a pilot you can honestly count. Nobody knows which jobs will survive, and this course does not pretend to.

So far there is only this table of contents. Lessons appear one at a time, and we promise no dates. Feed of new lessons (Atom)

17 lessons, 4 parts

Part 1 Tasks, not jobs

Economists have long counted the tasks inside jobs, not the jobs themselves, and that view is more useful than any list of dying occupations.

  1. A job as a bundle of tasks

    Computers take over routine tasks and complement the others, so a job changes even when its name stays.

    coming soon

  2. 47 percent or 9

    Two famous estimates of the risk to jobs differ fivefold, and the gap teaches you how to read any forecast about work.

    coming soon

  3. What machines take and what appears

    Displacement and new tasks run at the same time, and no law says they must balance.

    coming soon

  4. ATMs and bank tellers

    ATMs multiplied and bank tellers did not vanish: what really happened, and why the case is no general law.

    coming soon

Part 2 What has been measured

The data on generative AI at work is still thin and points in different directions, so it has to be read together.

  1. Who gains first

    In customer support AI helped newcomers most and barely helped the most experienced: what that may mean for you.

    coming soon

  2. Why paychecks barely show it

    Why saved time may not reach your earnings, and why that does not settle what comes next.

    coming soon

  3. The first rung

    What is known about hiring beginners in the jobs AI has touched most, and how to read such data.

    coming soon

Part 3 A small organization

A team of five has no adoption department: the rules and the counting fall to people already busy with their own work.

  1. Team rules on one page

    Which data may go where and who answers for the result: the agreements without which everyone decides alone and in silence.

    coming soon

  2. The first pilot

    One dull recurring task, one person responsible and a success criterion written in advance, in place of a big rollout across the firm.

    coming soon

  3. How to count whether it paid

    A measurement before you start, a comparison with the old way, and checking time in the costs, so the answer is a number, not an impression.

    coming soon

  4. It feels faster

    The feeling that work goes faster with AI can part ways with a measurement, which is why asking staff does not count as measuring.

    coming soon

  5. Replacing people too fast

    What happens when a company rushes to replace people to save money, and how to decide in advance what you will count as quality.

    coming soon

Part 4 Your next step

A plan for the coming year that does not depend on whose forecast turns out right.

  1. Take your week apart

    A plain table of tasks: what a machine already does, what it does with you checking, and what stays yours.

    coming soon

  2. Expertise as a bet

    A bold hypothesis: AI may widen the circle of people who can use expertise, given a foundation, so plan on it with care.

    coming soon

  3. A step sideways

    How to move into a neighboring role and take your knowledge of the field with you, and how to test the move cheaply before retraining from scratch.

    coming soon

  4. How to show what you can do

    Work people can look at, in place of “proficient in AI” on a résumé: what to gather, and how to say honestly which part the machine did.

    coming soon

  5. A look ahead: signals instead of forecasts

    Which numbers to watch in the next few years, from hiring of beginners to real use at work, and how to decide in advance what you will do if they move.

    coming soon