AI and Jobs

"Will AI replace jobs?" is the wrong size of question. Jobs are bundles of tasks, and tasks are what change. Some tasks in a role disappear, some are augmented, some become more valuable because they are now the ones that need a person, and some new tasks appear: supervising, evaluating, and designing agentic systems among them.

Decomposing roles into tasks is the practical exercise. It shows where augmentation is likely, where automation is plausible, and where the human contribution is the point. It also shows what new skills a role needs, which is more useful than a headline about replacement.

A word on the headlines. Much of what is said about AI and work on social media, and much of what is reported in the mainstream press, comes from people with agendas and interests. That includes this course, which is published by a company that builds agentic software. Learn to separate hype from fact. The facts are these. There is going to be disruption. Organisations, roles, and entire industries are going to be reconfigured. Many people will lose work or will have to acquire new skills to stay relevant, and it is already happening: freelance translators, for one, have seen the volume of available work fall sharply. At the same time, a language model is an elaborate next-word predictor going round in a loop. It is nowhere close to any kind of superintelligence, whatever the headlines say. Hold both of those facts at once.