AI is stealing our jobs. Why aren’t we stoked?

First, self-checkout came for the cashiers. 

But I did not speak up, because I am not a cashier.


Then autonomous driving came for the truckers.

But I did not speak up. Because if they don’t like it, they should learn to code. 


DALL-E and Midjourney came for the artists.

But that’s what happens when you don’t get a STEM degree.


Then AlphaCode came for the software engineers... 

…and there was no one left to speak for me.

-Poem inscribed on the first Employment Insurance claim ever submitted by a Computer Science graduate


Generative AI algorithms are so hot right now. Thanks to DALL-E and Midjourney, I can finally create art above a third-grade level. But if you read the news, you might not think such a benefit is worth the “stolen intellectual property” and “destruction of artistic integrity.” CHAT-GPT has an even worse publicist: Its applications seem limited to cooking up disinformation and cheating on essays.

This tension—between technology and the people it replaces—is inevitable. But history shows us that once robots or AI can offer a more economical or convenient experience, society will pivot in an instant. Want proof? Try getting a group of adults to order a pizza by phone. It’s like drawing straws for who gets the defective parachute. And no one is shedding tears over unemployed pizza-place-phone-operators.

Yet there’s no shortage of arguments against the use of image- and text-generating AIs. Most of them hinge on the fact that the AIs learn by ‘studying’ the works of human artists who never consented to their work being used this way. Meanwhile, the pro-AI crowd claims that this ‘studying’ is more akin to inspiration than plagiarism. 

These arguments are interesting. But they’re a complete and total distraction. Because even if artists could opt out of having their work included in AI training data, nothing would change. To wit:

  1. There are millions of illustrated works either in the public domain or owned by entities who would happily add them to training databases. This alone is enough to train sophisticated AIs. In fact, most AIs are trained on the massive LAION image dataset, where only three of the top 25 artists are even alive today. 

  2. In principle, AI does not need training data at all. AlphaGo is a beautiful example: Instead of being fed a dataset of human-played Go matches, it was simply given the rules of the game and then instructed to play millions of games against itself. It handily outperformed the best human players and all previous AIs. 

  3. Many artists are hung up on the fact that AI can replicate their style—often with remnants of a signature intact (the horror!). But for most commercial purposes, a unique, specific style isn’t necessary or even desired. Corporations don’t care enough about artistic innovation to sacrifice their bottom line; they just want an ad banner that looks good. Protecting unique work won’t stop businesses from using AI.

Therefore, no matter how much we whinge about ‘artist consent,’ the fact remains: Creativity is about to get democratized AF. And before long, AI will be better and more prolific than every human artist, writer, musician, video editor, accountant, and software engineer. The genie isn’t just out of the bottle; it’s dancing on the table and the floor is sticky with Dom Pérignon.

…So why doesn’t it feel like a party? 

***

The real problem

Look, I’ve been a writer for most of my life. Sometimes the work is fun and sometimes it’s even for a good cause. But mostly, it’s nothing more (or less!) than a day at the office. Let’s not pretend that writing a corporate video script or designing icons for a mobile app (a lightbulb, a bullseye, and a handshake, please) are the pinnacles of self-actualization or human creativity. They aren’t. And if AI can “make the logo bigger” and no one gets fired, everyone wins. Wouldn’t that be something?

What artists are really worried about isn’t plagiarism, consent, or whether AI art is ‘real’ or ‘takes effort.’ Like the rest of us, artists are worried about paying their rent. This isn’t paranoia. According to Ajeya Cotra, a senior analyst who studies A.I. risk, there’s a 15% chance AI will eliminate “most white-collar knowledge jobs” by 2036. 

So: Do we care? Definitely not yet and maybe not ever. So far, society has been mostly numb to the externalities of AI and automation. And when the paint of apathy chips, all we find behind it is an ugly wallpaper of platitudes and wishful thinking (“retraining!”, “learn to code!”). There’s no reason to think a swathe of unemployed software devs and graphic designers will jumpstart our bleeding hearts in a way displaced truckers and radiologists couldn’t.

What’s a society to do?

***

Finding a way forward

There isn’t much on the menu. First, we could ensure that all art used for AI training is in the public domain or used with the artist’s consent. This change would be welcomed with resounding applause. But it wouldn’t change much. And it certainly wouldn’t save any jobs.

The largest change would be to simply ban AI and automation or pass laws that prevent companies from replacing their flesh-and-blood employees. But think about what this is really saying: “AI does a faster, better job than you ever could. But we’re going to make you do it to earn your living.” This is the literal codification of busywork. And why are we banning the automation of creative jobs but not manual ones? Where are our tears for all the accountants obviated by Turbo Tax?

But what else can we do? Just let people live their lives!? 

Until now, the fundamental bottleneck of progress, growth, and innovation has been human effort. That era is ending. And at the risk of sounding like one of those Universal Basic Income (UBI) people, isn’t it fucked up that our response to the elimination of work is to worry about our bills instead of celebrating our emancipation? 

The fact that technology is taking our jobs isn’t a problem with technology. It’s the point. Most jobs suck. 

And art? AI couldn’t destroy that if it tried. To think otherwise isn’t just cynical—it’s to miss the point of human creativity. No matter how incredible AI becomes, we will still write and draw. We will still have trivia nights, spelling bees, hot dog-eating contests, and chess tournaments. We still do beautiful and ridiculous things like build cities in Minecraft and devote years to memorizing the first 100,000 digits of pi. We’re human. We can’t help it. 

So let’s not be so myopic that we dismiss our greatest achievements as a threat to the economy. Let's not be so closed-minded that we conflate the value of creativity with its value to the market. And let's not allow a vestigial attachment to “earning our keep” shut us off from the beautiful possibilities ahead.

The geyser of creativity is about to erupt. There’s no good reason for anyone to go thirsty.