Beware the workplace mouse trap

Life imitates art part 94. Scientists have discovered that lab mice may be conducting their own experiments on us. A paper published in the journal Current Biology and summarised here, speculates that mice seem to be testing their testers. They do this by deviating from simple expected behaviours such as responding to rewards to work out what might happen.

“These mice have a richer internal life than we probably give them credit for,” explained Kishore Kuchibhotla, senior study author and an assistant professor of neuroscience at Johns Hopkins University. “They are not just stimulus response machines. They may have things like strategies. Mice are surprisingly using higher-order approaches to learn even simple tasks, which may seem maladaptive. It may look like the animal is making a ton of errors, but during those errors, it’s actually getting smarter.”

To geeks of a certain age this might sound very familiar. That is because it is very like what happened with the mice who built the Earth in The Hitchhiker’s Guide to the Galaxy. Almost word for word like it, in fact. The realisation that the reporters of this story didn’t even allude to it makes me feel very old. All becomes clear, or unclear, at about 86 seconds.



The failure to attribute complex agency to the beings we observe and act upon is a common failure of our species. Take what is perhaps the most famous series of experiments in behavioural science, that of Stanley Milgram. In 1961, a team led by Milgram asked a number of people to administer what they believed to be increasingly high levels of electric shocks to a person in another room – in fact an actor – while listening to their responses and pleading.

The willingness of many to deliver what they thought were potentially fatal levels of electricity just because they were told to by an authority figure has cast a shadow on our understanding of people in the decades since.

But the results failed to take into account the idea that people knew or suspected it was an experiment on them, not the person in the other room.

According to an analysis of a previously suppressed follow-up questionnaire carried out by Milgram, it turns out that the people with the propensity to deliver the biggest shocks were aware that it wasn’t real.

Those who were reticent were those least likely to be aware they were the real subjects of the study. So the outcomes of the most famous experiment of its type are perhaps not what most people believe, even if it may be too late to challenge the laundered notions it created.

One of the inferences we can draw from this, apart from the way that an experimenter’s biases and presuppositions can distort their own research, is that people behave differently when they know they are being watched.

This is not just important to help us assess the outcomes of studies and research, but also because the act of observation can never be passive. Even in the era of isolation and homeworking, firms maintain the impulse to monitor people in real time, rather than work with and manage them in other ways. Firms have been almost as keen to acquire management software as they have collaborative tools.

This is something we will have to bear in mind as we enter a new era of observing people in the workplace and elsewhere. Our attitudes to perpetual embedded surveillance are doubtless shaped by our relationship with technology  so the pushback against the increasingly pervasive measurement of what we do at work is likely to remain moderate.

What we shouldn’t do is suppose that the people being measured are behaving in exactly the same way they would without scrutiny.


Of mice and men

The history of the humble computer mouse dates back to the 1960s and engineer Douglas Engelbart’s work on improving the way people and computers interact. He initially called the device he envisaged a ‘bug’ but the first prototype he created with Bill English was so unmistakeably a rodent that there was only one thing they could have called it. If only they had settled the question of whether the plural was mouses or mice.

In the intervening sixty years, the mouse may have lost its tail and ditched the mouse mat, but its basic design is largely unchanged, constrained by its core function and that of the human hand.

It has also become an unlikely weapon in the battle between remote workers and autocratic micro-managers. Since 2020, sales of productivity trackers have boomed. And with them, sales of devices that subvert the trackers, including something called generically a mouse jiggler.

As its name implies this is some sort of device that simulates mouse activity on a screen by moving the cursor. That keeps the computer active, prevents it going into sleep mode or closing apps. Its existence and popularity, of course, are direct consequences of the lack of trust between remote workers and their managers.

What gets measured gets managed cuts both ways, clearly.

There are probably bigger issues with this kind of management behaviour than remote workers gaming a system in this way. Perhaps most importantly, if you judge people’s performance on the amount of time they spend sitting in front of a computer, then that is what they will do (or pretend to do), whether it’s the right thing for the organisation or not.

It’s all nuts. If you have well-paid adults in senior positions reduced to buying devices that jiggle mice* to give the impression of work to other well-paid adults in senior positions, then something has gone very badly wrong. So too all those other trackers of eye movement, keystrokes and logins and whatever.

The very existence of mouse jigglers proves what happens when firms don’t trust the people they should trust. People game the system. They look for shortcuts. They focus on the wrong things. They repay mistrust with more mistrust. They act out in destructive ways. They lose faith in their employer and possibly their colleagues. And they aid in the development of a toxic work culture.

This sort of thing was foreshadowed in an old episode of The Simpsons. Looking for shortcuts as always, Homer is working from home in his role as nuclear plant safety officer but leaves his terminal with a drinking bird pressing the Y key to indicate “yes” to a series of questions and goes to do something more interesting. He returns to find that his bird has fallen over and a nuclear meltdown is imminent.

*mouses. Whatever.



Touch screens are ruining cars

AI startups are in big trouble

Recruiters are going analogue to fight the AI overload

How anxiety became a catchall for every unpleasant emotion

Europe is losing the productivity race