The latest workplace joke is a curious one: imagine taking leave towards the end of the month because your AI token limit has run out. There is no illness, family emergency, or holiday plan involved. Your digital assistant has simply stopped helping, and you are not sure how to carry on without it.
Is this actually happening at scale? There is no solid evidence yet that employees are routinely taking leave for this reason. However, the joke circulating on social media lands because it touches a real and growing issue regarding our dependence on artificial intelligence.
Artificial intelligence tools can save time, help organise ideas, explain difficult concepts, and take the drudgery out of repetitive tasks. For programmers, writers, analysts, and other professionals, they can be genuinely useful.
The trouble begins when assistance becomes dependence. If your productivity collapses the moment a tool becomes unavailable, it is worth asking how much of the work you can still do yourself. The corporate world is already confronting a related problem, with major firms facing questions around heavy AI usage, rising costs, and limits on how much employees can consume. The initial enthusiasm for AI is meeting the practical reality that its use is neither unlimited nor free.
At the same time, there is another cost that is much harder to measure. Look at the writing flooding the internet today. So much of it carries the exact same rhythm, carefully balanced arguments, predictable openings, and tidy conclusions. Despite having different subjects and authors, the output shares an oddly familiar voice.
Artificial intelligence is not solely responsible for this homogenization. Corporate editing, search-engine optimisation, and the pressure to publish quickly have encouraged formulaic writing for years. Generative AI, however, makes it remarkably easy to produce more of it at scale.
Good writing is not simply grammatically correct sentences arranged into neat paragraphs. It relies on unexpected observations, honest admissions, and specific details that only someone with direct experience would notice. It requires knowing when an argument needs more thought or when a sentence should be left alone.
These processes take time and often require sitting with a blank page, getting things wrong, and starting over. If every uncomfortable pause is filled by asking a chatbot for another draft, that creative process disappears, along with the opportunity to develop an individual voice.
The answer should not be to abandon AI altogether, but rather to stop treating it as a substitute for fundamental professional skills. Professionals can use these tools to challenge arguments, check structures, or spot gaps, but they must ultimately bring their own judgement to the work.
Writing without technological assistance on occasion and mastering professional fundamentals ensures workers can recognize when a machine is confidently delivering mediocrity. The token-limit joke is funny because it imagines a worker being forced to take a break when their digital crutch disappears, but the uncomfortable question remains whether many professionals are already approaching that threshold.