Am I Anti-Tech?
Or do I just have standards? Part 1.
When I was about 9 years old, my grandfather, Bill, bought a personal computer, and put it in his “office,” which was a closet in his garage. Bill was a lifelong tinkerer: he made his living at that time as a sort of industrial handyman, making sure HVAC systems were in good repair at large facilities. He could repair a full-size Ford truck, build an annex to his house (including doing the electrical work), and fix the small silver radio-controlled Lamborghini I got from Santa one year. Bill spent a lifetime accumulating gadgets and tools, which were overflowing from a storage unit by the time he passed a decade ago. And so it was in that tinkerer spirit he brought home that computer and invited me to tinker with it too.
The learning curve for a 9 year-old was steep with that computer, it not being a Windows PC or Macintosh with a graphical user interface but run on MS-DOS. I needed his help to boot it up and type in the right command to run the only “game” he had: Microsoft Flight Simulator. I’d be at my grandparents’ house in Mississippi over the summer, spending the days doing cannonballs into their pool and dribbling a basketball around in the dusty lot behind the barn where they’d put up a hoop for me. I’d spend the evenings in the garage on Flight Simulator, riveted by the experience of taking off in various pixelated small aircraft from now-defunct Meigs Field in downtown Chicago, doing laps around a pixelated Sears Tower, before trying to land the plane again.
When my dad bought a PC for our family a few years later — this one a sleek, black Acer with just-released Windows 95, a modem, and a CD-ROM that could play cutting edge games like Myst — my own tinkering escalated quickly, if only out of what counted as necessity for a 12 year-old. Our family Acer was a lemon; it rarely worked as it was supposed to. The process of installing games I wanted to play, like X-Wing, was fraught, and I spent as much time doing a primitive form of trial-and-error debugging as I installed and uninstalled as I did actually playing anything. This taught me patience and forced me to consider how the machine worked, and when my dad brought home books on HTML for Dummies and an intro to Visual Basic, I read them cover-to-cover, before coding my own Geocities website and my own turn-based game by the age of 13.
I’ve been called “tech-lashy” or “anti-tech” by peers. The line of thinking goes that if those of us who are working towards a better relationship between humans and the tech that’s being built today are consistently and uniformly too negative about tech companies and their products, that will be a turn-off both for the general public as well for potential allies (or, more importantly perhaps, donors) within the tech industry.
Such concerns most often come from technologists themselves; these folks might say that people like me just don’t understand because I don’t code. For someone like me with a degree in the humanities and whose profession is policy and politics, I couldn’t possibly comprehend how technology works, much less the limitations of what builders are dealing with, with the implication that expectations are too high and those of us with limited comprehension should just trust that the people on the inside are doing their best. One example is Casey Newton’s Platformer post from December, “The phony comforts of AI skepticism,” in which he argues that it tends to be outsiders who don’t understand who believe “AI is fake and sucks.”
I began this post with my own backstory because, even though I don’t know Python, I’m no tech troglodyte. Regardless of whether I think AI is fake (I do believe AGI hype is ultimately fake), I do think what we have sucks. But that’s not because I don’t understand; it’s because I’m experienced and knowledgeable enough to have standards. “The soft bigotry of low expectations,” a racially fraught phrase that entered the lexicon thanks to George W. Bush, comes to mind here as a way of explaining how we’ve come to accept that most digital products don’t work as they are promised to work. We’re doing ourselves a disservice the more we permit tech executives to escape dealing with the inadequacies in their products by simply gesturing at what might be possible in the future (if we just let them do whatever they want along the way and don’t regulate them, of course).
Consider, for a moment, the humble washing machine and dishwasher. When I was a Peace Corps Volunteer, I didn’t have hot water in my village house, much less any appliances, and so doing my laundry took all day and was physically exhausting. I first had to fetch enough water to do a load of clothes with — which is more water than is really a fun time to carry around — before using an actual washboard to scrub them. This was not a light scrub, mind you: you really had to work them on the washboard if you wanted to get them clean. (The babushkas in the village tended to have impressive Popeye forearms, I assume for this reason.) To wash the dishes, I had to first fetch water, then boil it, then had to do everything by hand.
When was the last time you used a washing machine or dishwasher that was unreliable? That had to be patched or updated? That was glitchy1? That needed to be replaced before it was 10 years old? Probably rarely, if ever.2 Washing machines and dishwashers tend not to be glitchy because the profit model of the companies that design and manufacture them is different from what has been the dominant paradigm in the tech industry for more than two decades and which has given us the “enshittification” of digital tech that we are forced to live with today. (Not to mention that the economic productivity gains of these appliances, which freed women from these sorts of household tasks and allowed them to enter the workforce, likely far outstrips the productivity gains provided by digital advertising and AI.)
Getting a washing machine or a dishwasher is not like adopting a Tamagotchi that has to be cared for incessantly — they aren’t technologies that ask for much or take anything from you, unlike, say, your average Microsoft product. Instead, these technologies give something to you; convenience, time, and a result that’s probably better and more sanitary than what you’d do yourself. We tend to exclude these sorts of machines from our conception of what we call “tech” these days, but the goal of a word processor or search engine or networking tool should be the same as a washing machine in the abstract: to more efficiently use human labor, by allocating as much of the necessary but low value-add labor to a machine that can actually do the job better than a human.
In that vision, Bosch and Whirlpool and LG and other home appliance manufacturers set out to design and make a machine that functions as it should all the time, and the companies compete with each other on fulfilling that promise of quality and on providing value to the customer. Importantly, if these companies make something that doesn’t live up to expectations or breaks, they are accountable both to customers and the market for that failure.
Yet none of these things are true of Google, Meta, Microsoft, or Amazon: these companies are so big that they are inescapable; realistic alternatives to their products don’t really exist because they’ve fully captured their markets; Section 230, the failure of Congress to pass any meaningful laws, and a weaponization of constitutional law deployed against state legislation have meant these companies are nearly impossible to sue; and most importantly, their profit model is built around collecting rents and extracting value from the customer, not delivering value.
So I am not anti-technology; rather, I am for technology that does a few things:
The technology should give exponentially more than it takes. Having to fiddle with something to get it to work through trial and error is one thing. Tinkering with early PCs taught me something, and generally learning to take care of your things is part of becoming an adult, so I’m not expecting zero work on my part here. But when Google’s search results are full of “One Weird Trick” ads and AI slop and it also is harvesting my data, their product is not satisfying this requirement.
Relatedly, the company making the technology should profit when it maximizes the product’s giving back to the human, rather than the inverse of profiting when it extracts or manipulates. A market that doesn’t create this incentive this is not allocating capital efficiently towards technologies that improve quality of life and productivity.
The productivity gains from the technology should be clear over the long-run, and not just substitute a different kind of work for the work it’s replacing. We are all our own travel agents and secretaries now, and we are also all busier than ever. That’s not a coincidence.
When maintained properly, tech should work 99% of the time as warranted, and the builder of the tech should make it right if not. We expect this of ATMs and airplane auto-pilot systems, and it’s totally reasonable to expect it of consumer and business digital tech. This includes not just a lack of glitches, but also fulfilling promises. To illustrate: my washing machine always cleans my clothes, and there’s a warranty if it doesn’t; evidence is mounting that Chromebooks in classrooms hinder educational achievement, for example, but with zero accountability for Google on the horizon for that failure. Interestingly, tech companies in start-up mode often make a big show of making it right if something went wrong — think of how good Uber or Amazon’s customer service was 15 years ago compared to today.
In part II, I’ll apply these rules to some existing products, share how I use generative AI, and discuss the policies that would incentivize more tech like washing machines rather than extractive and buggy tech we get.
As a remote worker — which is in many ways wonderful and in some ways not possible without the products I malign — I probably deal with at least one glitch a day, as I’m sure many of you do, if not more. Bluetooth headphones don’t connect. App needs an update and needs to reboot. Zoom keeps making me turn off the AI assistant. I could go on, and that doesn’t even cover the glitches that are there but we don’t notice, like when Meta’s algorithms flag and take down ads for women’s health products on Facebook because they are “sexual.” But I remember when I got my first bank card roughly 25 years ago being told to count the money you get out of the ATM in view of the security camera before you leave in case you get short-changed, and I can remember one time in my entire life that I’ve gotten the wrong amount of money — and it was because the ATM ran out of cash, not because of a glitch.
I recently had to purchase both of these appliances — closer to 15 years old than 10 — and I discovered it’s surprisingly hard to find a dishwasher or washing machine that doesn’t have WiFi in it, which obviously complicates the machine and makes it more likely to be buggy.