Word Processing, AI, and other new things we struggled to define
A better way to think about the future of work
By now, you will have figured out that I think our present moment parallels the software revolution that transformed office work beginning in the 1970s. By the middle of the 1990s, a new paradigm for work had emerged around the modern, boring office suite. That paradigm remained pretty stable until March of 2020.
COVID is actually useful exception to prove the rule of that stability. Going remote was possible because of the software revolution. Much of our work had consolidated into a single computing device that could be sent home. When things went remote, people essentially looked for 1:1 digital surrogates for most work. The nature of the work itself did not fundamentally change in spite of that massive upheaval.
Pretty much anyone who has worked with generative AI thinks we are on the precipice of a paradigm shift in how we work. A new paradigm is incredibly difficult to imagine, harder to describe. We experienced a paradigm shift more recently, but not so much in how we work. Instead, social media brought a shift in how technology mediates the world to us. As I wrote previously, I think we are caught fighting the last war if we think of the present transformation as a continuation of that paradigm shift.
AI will further transform how the world is mediated to us, of course. Still, I think the impact on work will be more profound. Also, general office work has been more resilient to major changes over time than media, which has changed pretty regularly over the past century.
If we want to think strategically right now, we should be thinking about the last time work really changed. We should be talking about revolutionary and fundamentally boring technologies like word processing.
Typing a document has become perhaps the most intuitive thing we could imagine doing with a computer. This video from 1975, however, demonstrates that that task was neither easy to imagine nor to describe, not even for people who typed documents all day, every day.
To me, this video seems entirely alien and eerily familiar. It feels a little too on the nose, almost like Saturday Night Live decided to lampoon the way we talk about AI right now by creating a period sketch in which people talked breathlessly and nonsensically about a revolutionary new technology we now regard as mundane.
By the way, a sketch comedy show did do that. Back in the early days of office computing a Norwegian television program lampooned hapless users of computers and the IT professionals who staffed the help desk, imagining the chaos brought on by the new technology of the book.
But the video above is not a joke from the present, it is a real window into the past. This video provides a candid account of the gendered assumptions people made about work in 1975. The way work was gendered in that period really mattered and it matters how we gender work today, even if we are much less explicit about it.
Remember that when computers as computers began crunching numbers decades before this, they had replaced “human computers,” a gendered job performed almost exclusively by women on behalf of the mostly male engineers. When computers came to the office, they augmented or automated women’s work, so women learned to operate them.
The intersection between gendered work and technology deserves far more attention than I can give it here, so I will have to return to this in the future. But, if we are thinking about applying computing to specific kinds of office work, it really matters that men and women in that period did very different kinds of work. It matters that men and women understood different workflows, but also the same workflows from different angles. It matters just as much that women’s time and labor was generally valued less in every sense.
For the moment, can we appreciate the overly technical word salad offered here as a plain definition of word processing? “Word processing,” the narrator explains, “is a way to transform ideas into written communication, quickly and accurately, through the use of automated equipment operated by skilled people using revised office procedures.” Oof.
The word processing software described in this video is not what we now use. This software was more like a benign application of a key logger placed in an electronic typewriter, slowly piecing together an invisible, final record of keystrokes that could eventually be printed. There, I struggled to define it. It really is difficult. Or, as the narrator says, “spooky.” If you are really curious, many of these text editors are still accessible via the command line on your computer. Go try them out and experience the joy of trying to exit vim.
I am still working on this idea, but it strikes me how much more work became self-service and, therefore, individualistic in our familiar paradigm compared to what existed previously. A different advertisement for the AXXA 90 system—about which I can find almost nothing—anticipates a lot of the language we use for AI, but it also drives home the newness of something we now take for granted: We expect to be able to look up any information we need, whenever we want it, on our own.
This is remarkable because it shows executives using computers to take on work that had previously been delegated. I am curious what readers think about this, especially if anyone actually experienced this earlier era, which significantly predates me.
If we struggle to understand these videos, it is because we will not necessarily understand the assumptions they made about work in that moment, about how tasks fit together and how they might come apart when a computer could do some—but not all—of the parts of what had always seemed like a single task. It is easy for us to understand how powerful shared electronic calendars or electronic documents are; it is much more difficult to understand how people worked without them, the assumptions they made that we have forgotten that made all of this a little “spooky.”
Perhaps we should be a little nicer to ourselves as we squint into the future, hyping future applications of technologies we barely understand. It is hard to imagine a paradigm shift simply because it is.