How is digitization changing work? An introduction

Does the "coder-king" increasingly rule in all sectors? Not necessarily; users often push back. And often, coders are working to implement someone else's ideas.

Made by Google Gemini when I gave it the headline and the tagline.

Clearly, the digitization of life — the fact that we use networked software for almost every task — means that the way we work has changed. We have managed to embed software even into tasks that would seem to be undigitizable: truck drivers drive machines that are peppered with sensors which feed into “analytics” programs that can spit out metrics of performance that their bosses can then use to evaluate them. Even our taxi-drivers, because they access jobs through apps like Uber and Lyft, need to adopt different practices to accommodate the software. Clearly, the digitization of work matters and has changed what workers across the board do.

But I think analysts — especially in journalism and those in my corner of the social sciences — are too quick to jump from that to a mostly dystopian conclusion that coders rule us. Here, for instance, are my colleagues Jenna Burrell and Marion Fourcade in their survey article “The Society of Algorithms” (already cited more than 500 times, according to Google):

In its quest for market expansion, the tech industry increasingly carves away at and lays claim to tasks that once were protected as the proper domain of professional judgement, in every occupation from business management, medicine, and the criminal justice system, to national defense, education and social welfare. Deeply entrenched jurisdictions and human institutions are nothing but transient compromises awaiting the automation of their most routine tasks, and sometimes their entire work domain. And no profession, no matter how prestigious or how high the barriers to entry, is exempt from having its judgement subject to a second (algorithmic) opinion, if not wholly supplanted by it. Legitimacy has been displaced from the professional to the coder-king –and, increasingly, to the algorithm. [my emphasis]

You should read the whole thing because it synthesizes a lot of interesting work, aims to provoke, and is written with verve and style. But it is also shot through with this apocalyptic tone, exemplified in that paragraph I quoted, that I find somewhat unnecessary. The rationalization of work has a long history that precedes the computer and it’s not clear to me that the “tech industry” has any special power to disrupt “deeply entrenched jurisdictions.” But even more important, what’s wrong if some deeply entrenched jurisdictions are disrupted? Generally, interpretive social scientists have regarded deeply entrenched professional jurisdictions as a matter of social control and looked upon them with suspicion. But it seems like today, with professionals being forced to work with software, social scientists would rather defend professional jurisdictions rather than side with the “tech industry.”

On the level of mechanisms, though, paragraphs like the one I quoted use terms that can mean a whole lot of things, e.g., “tech company,” “professional judgement,” “algorithmic opinion,” and so on. Is a long established company that makes machine learning software that will monitor infants in a NICU a part of the “tech industry”? Or does a company have to be venture funded to qualify for that? Or is it that they need to make technology that incorporates a lot of machine learning? Who are the “professionals”: are we using the traditional definition that includes lawyers and doctors and social workers but excludes truck drivers? Are these “algorithmic opinions” binding on professionals or are they merely supposed to be guidelines?

Over the years, as this has become the dominant view in the interpretive social sciences — that tech companies are hungry to automate and seek to remove every vestige of human judgement in decision-making — I have tried to read as many empirical studies as I can about what is actually going on with digitization and workers. There are many such studies out there and what I have sampled is barely the tip of the iceberg. But what those studies show is that not every workplace is the same; that, yes, while some workers are losing control over what they do, others are able to retain it. And still others are finding their jobs transformed without necessarily losing control over what they do; when they do lose control, it’s not to the “algorithm” but rather to their audience.

So what I’m hoping to do on this blog is write a series where I focus on one or two workplaces and occupations in every post and document exactly how work is changing in response to digitization and what this means for workers. Some of the occupations I am thinking about include:

• Policing

• Courts (judges)

• Lawyering

• Social work

• Journalism

• Content creation

• Trucking

• Driving (taxis)

• Medical work (doctors, nurses)

• Marketing

• Political consulting

• Education (how teachers teach and students learn)

• Customer service.

I may also explore science: how physicists, biologists, and others are handling the deluge of data. I am sure still others will occur to me as I write. If there are occupations you think would be interesting to explore, please leave a note in the comments (and even better if you already know a great book or article that explores how that work is actually done in practice).

When I write these posts, my goal will be to focus on the way the software is designed and how it is used and the interaction between the two.

When it comes to design, it bears mentioning that while software obviously needs to be coded, programmers and coders as such may play a very small role in thinking about what features to build and/or emphasize. The people who build software engage in a number of activities including talking to future customers about their needs and then figuring out which of those needs are best addressed given the resources the developers have available. Understanding this process tells us about what values are embedded into technical systems and why.

But design only works if users use it the way it’s been intended. And that isn’t always the case. Sociologists Sarah Brayne and Claire Maiers, for instance, found that police and NICU clinicians absolutely did not use the digital systems in the ways the designers of those systems intended. How workers use these systems really depends on factors ranging from work culture to professional autonomy and it is important to explore these in detail.

I think that’s enough of an introduction!

In my next post in this series, given the recent discourse on journalists and influencers and what norms are (and should be) appropriate for each, I’ll start with those two occupations: journalism and content creation.