The rise of remote software teams and the impact of A.I.

‍

This was the year when the impact of AI on society – a long-anticipated revolution – became something much, much bigger. It was the Industrial Revolution. It was the Space Race. It was another Internet boom. It was all of those together. AI is too powerful, too opaque, too dystopian, too dangerous. And now, at last technology has gone too far. After all, this introduction (parts in bold) were drafted by A.I. (ChatGPT 4) with the help of one of Ross Douthat’s Atlantic articles. However, the consequences of similar technological movements point us to ramifications we can expect to see with A.I. The disenfranchisement, disruption, and dystopian promises of A.I. aren’t new.

‍

I remember back in 2016 when I first had a schedule that allowed me to work from home (WFH). It was exhilarating to be able to host virtual meetings from my apartment and squeeze runs in between calls. I would “Zoom” with my Australian clients who were in the height of their summer while I was watching Colorado winter storms from the window. I felt as if I were Oz behind the curtains, pulling strings and synchronizing my team of project managers and developers without even needing to don a pair of shoes. 

‍

Then, in 2020, COVID greatly democratized the number of people who were able to work from their own hearth, allowing for almost anyone who used a computer to sign in remotely. The lockdowns accelerated the development of technologies needed to support virtual industries. Server farms, APIs (databases talking to databases), and communication tools grew and found their metamorphosis in real-time video conferencing software (Zoom, Teams, etc.) as well as a bevy of other tools. These new tools promised the same things that A.I. now promises: work (and life) reimagined without friction. You could build new software, conduct meetings, and do “team building” with a few clicks, all while wrapped in the warm embrace of your pajamas. Sounds good, right? 

‍

However, COVID was only the accelerant. Economic forces primed the development long before we had anything called “Slack,” “AWS,” or any other app or technology that preceded A.I.. One example of this is offshore software development teams. Over the past few decades software consultancies and tech giants have seen ripe fruit in the growing population of cheap technical labor in places like India, Pakistan, and eastern Europe. And it’s no surprise why; these developers can cost as low as 10 cents on the dollar compared to what a U.S. developer would cost. Unlike the more recent trends to bring back U.S. manufacturing jobs or to shop local software development companies are no longer searching for, “software developers near me.” 

‍

We’ve all rubbed shoulders with this trend via the overseas call center you’re connected with when you want to change your ISP or get technical support for a piece of hardware. Still, the same places aren’t just picking up our calls; they are also building the software that’s on our phones, smart watches, and social media feeds. It’s Adam Smith’s invisible hand simply matching labor to markets at the right price. It’s Henry Ford’s assembly line transcended, efficiently connecting the building blocks of the software development lifecycle. We have our product managers and executives in Silicon Valley (or now completely remote) and our engineers in Bangalore, Karachi, and Kiev all connected through emojis, avatars, Slack messages, and zoom calls. No A.I. in this mix yet, but the same dystopian promise of a world of perfect efficiency. 

‍

A consequence of twin revolutions of market forces and technology hasn’t been mass unemployment or the data privacy concerns pundits and economists promised. Sure both of these things have happened at some scale, as seen in the current congressional bout with TikTok and the 2023 tech layoffs. In time, Congress will reach a deal with TikTok or ban it from U.S. markets, while allowing for countless others to fill its place. Tech workers will adapt, learn a new set of tools, and pick up roles doing slightly different things. So, what’s the more permanent consequence we’re seeing? It’s one that seems to be what we’ve been told is the main advantage: an ecosystem of technology that frictionlessly connects distributed parties. Software developers are connected to product managers, customers to tech support, and finally all the power of A.I. to anyone with a wireless connection.  

‍

When I first went to a grocery store in New England, about two years ago, I was startled by the experience. No greeting, no eye contact, no words, nothing. Did I have a budding zit on my forehead? Did my unibrow come back? What gives? Sure, ALDI prides itself on its prices, not its customer service, but still. What I came to find out in the weeks following is almost all of my interactions at supermarkets in my new home were about the same. I'm sure part of it is the particular city I’m in, situated between the two large metropolises of Boston and New York. The larger part of it, I would argue, is a regional culture that’s been designed for frictionless interactions between customer and business. It’s the TSA effect, where everyone goes through the same line, follows the same rules and is dealt with in the least amount of friction possible. You give me your groceries, I’ll check you out, and we don’t need to talk or interact. 

‍

Unlike passing through TSA or buying a tin of Pepperidge Farm Pirouette Cookies at ALDI, software development doesn't require you to show your face or engage in any human-to-human interaction, even if it's a sterile one. Thus, the tendency to decouple the person from the work is so much more ubiquitous. This is also why Hollywood has had so much fun with the image of the basement-dwelling hacker breaking into a central government agency and stealing all their secrets. The developer doesn’t need to be in person to do any part of the job. 

‍

A.I. will continue the trend of decoupling the human from the action on the promise of removing tedium and improving efficiency. It won’t replace us; it will just help us forget who we are – namely humans – who can only delude ourselves into thinking we can escape our physicality. Interactions will be replaced by a process or tool. I will write a product requirement (aided by AI), then I will send that requirement to a developer via a chat tool, then the developer will do the work (greatly aided by A.I.) and when it’s done, it will be automatically marketed as finished. Bare minimum human communication or creativity will be required with each of us working from our own corners of the world, isolated, but operating with maximum efficiency. 

‍

But this automated progress isn’t inevitable.

‍

PART 2 TO COME.

‍