Should Your Boss Manage Your Emotions?

 

The increase in telework since 2020 has led to a corresponding increase in so-called "bossware," the type of software that tracks employees' behavior. Some of this just reflects a tendency to micromanage, like the exhortation "cameras on, people!" during a Zoom meeting (were they really paying any more attention when you met with them in person?). Some new products introduced in the last 5 years treat knowledge workers like assembly-line workers, rating their "productivity" in terms of mouse clicks and documents opened. Again, this doesn't seem like it will actually increase work completion -- Charlie Warzel and Anne Helen Petersen's book Out of Office argues that it instead probably just increases "performative work" or "LARPing your job" [LARP = "live action role-playing," for my less-geeky friends and colleagues out there]. It also feels pretty intrusive to people who are used to managing their own schedules and task lists. Where does workplace monitoring cross the line from professional to overly personal and intrusive?

Here's a relatively benign example of the type of software that I'm thinking of, which you have likely seen if you're in a work environment with Microsoft Office products: 

This graph tells me that I get invited to far too many meetings: About a third of them overlap with other meetings, less than half have high attendance (a sign that too many extraneous people were invited), 17% of them lasted more than an hour (way too long), and I multitasked during nearly half of them (mea culpa). On the other hand, my college is a pretty orderly place -- all of the meetings had advanced notice, occurred during working hours, and ended on time, and I RSVP'ed to all of the meeting invitations in advance. The statistics also show that of 46 meetings in the past month, I only organized two of them -- so mainly it's other people's meetings that I'm attending. Now, I'm in a middle-management role, so some of this is to be expected. Yet it also provides some quantitative support for my sense that I spend too much time in too many meetings that could have been an email. 

So meeting tracking is all right with me. MS Outlook also gives me statistics on how many other workers responded to my emails after hours, which has given me an incentive to stop sending so many after-hours emails (I still write them when I have the time, which is often in the evening, but I now use "scheduled delivery" to send them first thing on the next workday instead). Outlook even rewards me by telling me how many people I saved from reading my annoying after-hours messages! Overall, this might contribute to a saner workplace, where people actually get to have time free from work.

I get a little more irritable when Outlook starts to criticize my own all-hours work habit, like this: 

This image shows that on 22 out of 30 days, I was working at times of the day when Outlook thinks that I "shouldn't" have been working (and that despite the fact that I have already tweaked my "office hours" to include a generous 12-hour window for potential work). This "quiet time" effort mostly involved writing email responses at night (because I spent my days in meetings!). However, another graph shows that the emails weren't actually sent at night for the most part, so my scheduled-send strategy does seem to be working. Nevertheless, Outlook habitually gives me a guilt trip for not taking better care of myself in the off hours. This annoys me in a way that the reminder not to disturb other people's off hours does not: Shouldn't I be the judge of whether getting an email off my plate relaxes me more than, say, watching TV?

The Outlook developer might respond that they are simply giving me some information, as with the meeting statistics, that I could find useful in managing my own work life -- it's up to me what to do about it. But it's also true that the graphical aspects of the interface shown above match common gamification approaches to behavior change, such as the use of streaks or badges. Simply knowing that I am being monitored is also likely to have an effect on my behavior. The repeated emails about my work habits create a compelling narrative about how I "should" be working. The net impact is to shape my work behavior toward some pre-determined ideal, using many of the same tools that B. F. Skinner employed when teaching pigeons to play ping-pong. 

Here's another example, showing a heads-up display that provides real-time feedback to telehealth workers on their word choice, tone of voice, etc. -- in this case based only on the health care worker's verbalizations, but in the future maybe by monitoring the other person's reactions as well. Note the yellow dot in the 4-quadrant system at the lower left, together with some keywords that may reflect the patient's emotional state.

This example is a training simulator designed to help health care professionals learn to interact more effectively, which is all well and good (click here if you're interested in signing up for their training). But as AI methods improve, one could imagine this type of real-time display being applied in actual practice, and even that telehealth employees might be graded or reimbursed based on their performance as determined by an algorithm. Microsoft recently announced a "speaker coach" addition to Microsoft Teams, which will provide feedback immediately after you give a presentation, including information about your pacing, level of repetitiveness, and use of culturally sensitive words (more info here).

Philosopher Eric Schwitzgebel envisions a world in which even more direct means (chemical or neuroelectrical) might be used to directly affect employees' performance. He writes:

I don't know to what extent people in Silicon Valley, Wall Street, and elite universities already use high-tech drugs to enhance alertness, energy, and concentration at work. What I'm imagining is just a few more steps down this road ... . If Starbucks can select between applicant A who is willing to crank up the perky-friendly dial and applicant B who is not so willing, they will presumably prefer applicant A. ... If Stanford wants to hire a medical researcher who publishes high-profile studies at an astounding rate, they'll likely discover that it's someone who has dialed up their appetite for work and dialed down everything else.

Adrian Hon, in his book A New History of the Future in 100 Objects, similarly envisions a near-future where humans and AI applications together form an "amplified team" for high-performance work, as in this description supposedly written by a company that supports this type of team:

The Basics. Amplified teams contain 3 to 7 human members supported by software that allows them to communicate with each other and with AI support systems at an accelerated rate. Unlike traditional networked teams, most amplified teams work in close physical proximity in order to reduce lag and take advantage of nonverbal cues that increase trust and bandwidth. This helps develop heightened levels of team reasoning and "limited collective utility" maximization. ...

Do I Still Keep My Personality? Of course! While some amplified teams choose to present a "united" face to the public, even the most integrated team members remain fully capable of an independent life. It is true that after leaving a team, some detached members may appear to have a substantially changed personality, but the same would be true of anyone undergoing a significant life event. At Amplified Green, we pride ourselves on our detachment expertise, with a 99.6 percent success rate as measured by the International Amplified Team Organization.

What Kind of Linking Technologies do Amplified Teams Use? Although magnetoencephalography and dermal tattoos are currently in favor, some teams have had success using extremely low-tech solutions such as necklaces and glasses, or even terminals. Others prefer to use experimental and even physically invasive techniques. Unless all of a team's members are experienced and confident in a particular linking technology, we usually recommend that they try out the full range before settling on a solution. ...

Are Amplified Teams Good for Society? Though they are still comparatively new, amplified teams have made countless positive contributions to society. They have made technical and scientific breakthroughs, improved the efficiency of companies, charities, and governments, and produced critically and popularly acclaimed works of art. While some claim that teams are displacing other workers, the truth is more complicated and cannot be measured purely in employment figures and inequality indices. Many advances made by teams have created new jobs and new forms of wealth, and many teams voluntarily fund retraining programs for anyone affected by their work.

This dystopian future suggests a loss not just of conscious control over one's work-life decisions, but maybe also of one's individuality. 

How close are we to this world already? I have seen a tool (not at my university) that listens in on an employee's phone conversation and runs their tone of voice through an AI pattern-recognition algorithm. If the AI detects evidence of anger or frustration, it sends the employee a message saying "you seemed upset during that last phone call. Would you like to talk to someone about it?" The message also includes a direct link to the company's contracted mental health support provider. This is not a hypothetical like the training tool shown above -- it is a real-world product currently on the market. Warzel and Petersen give another example in which a company listened in on conversations via digital ID badges, to track "who is talking to whom, for how long, with what tone of voice, how quickly they speak or when they interrupt, etc. to try to identify what makes for a good team" (p. 176). Both of these examples feel highly intrusive to me, perhaps because they involve listening in on conversations and attempting to draw inferences about someone's private mental state. 

What if we don't want to use these tools? Professor Schwitzgebel suggests that 

At first, employees will probably keep the hand on the dial themselves or mix the drug cocktails themselves. ... The employer might provide an initially voluntary "energy drink" for all employees at the beginning of the shift. High-status employees, in contrast, might more effectively keep their own hands on the dials. However, the pressure, and consequently the indirect control, might be even more extreme among elite achievers. If mood interventions are highly effective, then they will correlate highly with professional performance, so that as a practical matter those who don't dial themselves to nearly ideal settings for work performance will be unlikely to win the top jobs.

In other words, unfettered capitalism will continue to propel the march toward more and more intrusive technologies of monitoring and control. Warzel and Petersen draw much the same conclusion. They write that until we start to view workers as what they are, a collection of human beings, we will only make things worse for them (and in the long term, worse for actual productivity) by applying new and more complex mechanisms for worker surveillance and behavior modification.