Workers across wildly different jobs seem to use the same test for whether a task is pointless: can they tell who it actually serves?

This explores how workers judge whether a task matters or is pointless, and whether that judgment depends on their occupation. The corpus doesn't have occupation-by-occupation definitions of meaningful work. What it does have suggests a surprising answer: the line between meaningful and pointless is drawn at the level of the task, not the job title. One study asked 202 workers to rate 171 tasks on a five-item 'bullshitness' scale. The tasks that came out as pointless were the ones that felt disconnected from what the organization was actually trying to achieve. That pattern held across 22 occupations and 12 sectors Does perceived task meaninglessness drive worker desire for AI delegation?. A nurse and a software engineer may do very different work, but they seem to spot pointless work the same way: they can't see who or what it serves.

That judgment has practical effects. The more pointless a task feels, the more willing workers are to hand it to AI, and the less they want to check its output. Perceived meaning also changes what kind of AI helper people want. For pointless tasks, workers prefer agents that are fast, simple and decisive over ones that explain themselves or get creative What agent traits do workers prefer for meaningless tasks?. So 'what do I want from AI at work?' has no single answer. It depends on whether the worker cares about the task in front of them.

Zoom out from tasks to whole jobs and a related idea appears. Romero argues that the jobs safest from automation are not the most complex ones. They are the ones where people pay for a particular person rather than for an output Which jobs will actually survive automation by AI?. Read next to the bullshitness research, this hints that meaning tends to sit where work connects to a goal or a relationship someone actually values. Where people really do delegate to AI fits this loosely: real handoffs cluster in information-heavy work and follow what the technology can do, not old predictions about which 'routine' jobs would go first Where have workers actually delegated tasks to AI?.

The twist you might not expect is that AI can create new pointless work. Desk workers estimate that 15.4% of the work they receive is unhelpful AI-generated content, and that each instance takes almost two hours to deal with. That is longer than if the sender had just done the work themselves How much work that employees receive is actually unhelpful AI content?. More broadly, AI tends to move time around rather than save it: time shifts from doing the task to writing prompts and checking outputs Does AI really save time, or just change how we spend it?. A large Microsoft trial found Copilot cut email time without changing what people actually worked on Does Copilot change what knowledge workers actually do?. Together, these findings suggest that handing pointless tasks to AI doesn't automatically free people for meaningful work. Sometimes it just produces a new kind of pointless task.