Nothing ventured, nothing changed

Think of something you have been meaning to switch to and haven’t. A different bank, a different phone plan, a different way of running your Tuesday mornings. You know the alternative is probably better. Now notice how quickly the reasons for not having done it arrive: the setup, the forms, the fiddling, the chance it turns out worse. Those reasons are not excuses. They are real costs, and they arrived faster and clearer than the benefits did. And now the trickier part: see if you can bring to mind something you have been doing for a long time and have not considered changing. It is difficult even to generate an answer, let alone to ask whether it is optimal, and that difficulty lives in the non-choice-ness of the experience, the feeling of merely carrying on with what already is rather than choosing it. That feeling has a name, status quo bias, and while this piece of cognitive machinery has large effects on our lives, we mostly remain oblivious to its workings precisely because of that non-choice-ness.

What this means is that a failure to switch gears often cannot be explained by ignorance of the benefits. Many people already understand there are benefits to using AI tools at work. In a survey of 767 knowledge workers, 68% believed AI tools would improve their job performance, while only 37% used them regularly (Berman et al., 2025).

Of course, some of that gap can be explained by real reasoned positions, for instance, when workers believe their organisation lacks the infrastructure to support meaningful adoption (e.g., Zhu & Gao, 2026). However, reasoning rarely explains the largest chunk of inertia. What accounts for much of it is stranger and harder to see, because it rests on how choices are shaped rather than on what people believe.

When a choice is framed as the default, it is far more likely to be chosen (Samuelson & Zeckhauser, 1988). For instance, when two American states offered drivers the same two car insurance options but made different ones the default, the majority in each state kept whichever one they had been given (Kahneman et al., 1991). This reluctance to switch is a powerful finding within the psychology literature, and its presence has been demonstrated across a wide range of domains, from things as trivial as swapping a pen (Kahneman et al., 1991) or a coin (Gal & Rucker, 2018) to potentially high-impact decisions like choosing jobs or where to invest money (Samuelson & Zeckhauser, 1988; Xiao et al., 2021).

Why the pull? One long-standing explanation is loss aversion: what we give up weighs more than an equivalent gain, so the losses of moving loom larger than the benefits of arriving (Kahneman et al., 1991). Consider what is actually at stake in a working week. In economics the currency is money. At work it is time, and the costs decompose into three: the hours to learn the new thing, uncertainty about whether it will work, and the value of skills in the current way of working that switching appears to write off. Across an enterprise system rollout, these switching costs were the central determinant of resistance, and they mediated nearly everything else, including how capable people felt of adapting to the new system and whether they believed their colleagues were in favour of the change (Kim & Kankanhalli, 2009). The third cost is the quiet one. Someone who has spent four years becoming fast at a task is not being irrational when a tool that makes them slow for a fortnight feels like a loss. It is a loss. Whether it remains one over a longer horizon is another question entirely.

The trickiest part of this whole cognitive system is its quiet voice. A default option barely registers as a conscious choice, if it registers at all, and the heaviness of losses relative to potential gains is likewise an automatic instinct that often evades conscious detection. This means that long familiarity with the current way of working changes how easy and how advantageous the new option appears, before any weighing happens (Polites & Karahanna, 2012). The scales are adjusted before anything is placed on them.