How to fight AI at work

Your Mileage May Vary is an advice column offering you a unique framework for thinking through your moral dilemmas. It’s based on value pluralism, the idea that each of us has multiple values that are equally valid but that often conflict with each other. To submit a question, fill out this anonymous form. Here’s this week’s question from a reader, condensed and edited for clarity.

I’m an AI engineer working at a medium-sized ad agency, mostly on non-generative machine learning models (think ad performance prediction, not ad creation). Lately, it feels like people, specifically senior and mid-level managers who do not have engineering experience, are pushing the adoption and development of various AI tools. Honestly, it feels like an unthinking melee.

I consider myself a conscientious objector to the use of AI, especially generative AI; I’m not fully opposed to it, but I constantly ask who actually benefits from the application of AI and what its financial, human, and environmental costs are beyond what is right in front of our noses. Yet, as a rank-and-file employee, I find myself with no real avenue to relay those concerns to people who have actual power to decide. Worse, I feel that even voicing such concerns, admittedly running against the almost blind optimism that I assume affects most marketing companies, is turning me into a pariah in my own workplace.

So my question is this: Considering the difficulty of finding good jobs in AI, is it “worth it” trying to encourage critical AI use in my company, or should I tone it down if only to keep paying the bills?

Dear Conscientious Objector,

You’re definitely not alone in hating the uncritical rollout of generative AI. Lots of people hate it, from artists, to coders, to students. I bet there are people in your own company who hate it, too.

But they’re not speaking up — and, of course, there’s a reason for that: They’re afraid to lose their jobs.

Honestly, it’s a fair concern. And it’s the reason why I’m not going to advise you to stick your neck out and fight this crusade alone. If you as an individual object to your company’s AI use, you become legible to the company as a “problem” employee. There could be consequences to that, and I don’t want to see you lose your paycheck.

But I also don’t want to see you lose your moral integrity. You’re absolutely right to constantly ask who actually benefits from the unthinking application of AI and whether the benefits outweigh the costs.

So, I think you should fight for what you believe in — but fight as part of a collective. The real question here is not, “Should you voice your concerns about AI or stay quiet?” It’s, “How can you build solidarity with others who want to be part of a resistance movement with you?” Teaming up is both safer for you as an employee and more likely to have an impact.

“The most important thing an individual can do is be somewhat less of an individual,” the environmentalist Bill McKibben once said. “Join together with others in movements large enough to have some chance at changing those political and economic ground rules that keep us locked on this current path.”

Now, you know what word I’m about to say next, right? Unionize. If your workplace can be organized, that’ll be a key strategy for allowing you to fight AI policies you disagree with.

If you need a bit of inspiration, look at what some labor unions have already achieved — from the Writers Guild of America, which won important protections around AI for Hollywood writers, to the Service Employees International Union, which negotiated with Pennsylvania’s governor to create a worker board overseeing the implementation of generative AI in government services. Meanwhile, this year saw thousands of nurses marching in the streets as National Nurses United pushed for the right to determine how AI does and doesn’t get used in patient interactions.

“There’s a whole range of different examples where unions have been able to really be on the front foot in setting the terms for how AI gets used — and whether it gets used at all,” Sarah Myers West, co-executive director of the AI Now Institute, told me recently.

If it’s too hard to get a union off the ground at your workplace, there are plenty of organizations you can join forces with. Check out the Algorithmic Justice League or Fight for the Future, which push for equitable and accountable tech. There are also grassroots groups like Stop Gen AI, which aims to organize both a resistance movement and a mutual aid program to help those who’ve lost work due to the AI rollout.

You can also consider hyperlocal efforts, which have the benefit of creating community. One of the big ways those are showing up right now is in the fight against the massive buildout of energy-hungry data centers meant to power the AI boom.

“It’s where we have seen many people fighting back in their communities — and winning,” Myers West told me. “They’re fighting on behalf of their own communities, and working collectively and strategically to say, ‘We’re being handed a really raw deal here. And if you [the companies] are going to accrue all the benefits from this technology, you need to be accountable to the people on whom it’s being used.’”

Already, local activists have blocked or delayed $64 billion worth of data center projects across the US, according to a study by Data Center Watch, a project run by AI research firm 10a Labs.

Yes, some of those data centers may eventually get built anyway. Yes, fighting the uncritical adoption of AI can sometimes feel like you’re up against an undefeatable behemoth. But it helps to preempt discouragement if you take a step back to think about what it really looks like when social change is happening.

In a new book, Somebody Should Do Something, three philosophers — Michael Brownstein, Alex Madva, and Daniel Kelly — show how anyone can help create social change. The key, they argue, is to realize that when we join forces with others, our actions can lead to butterfly effects:

Minor actions can set off cascades that lead, in a surprisingly short time, to major structural outcomes. This reflects a general feature of complex systems. Causal effects in such systems don’t always build on each other in a smooth or continuous way. Sometimes they build nonlinearly, allowing seemingly small events to produce disproportionately large changes.

The authors explain that, because society is a complex system, your actions aren’t a meaningless “drop in the bucket.” Adding water to a bucket is linear; each drop has equal impact. Complex systems behave more like heating water: Not every degree has the same effect, and the shift from 99°C to 100°C crosses a tipping point that triggers a phase change.

We all know the boiling point of water, but we don’t know the tipping point for changes in the social world. That means it’s going to be hard for you to tell, at any given moment, how close you are to creating a cascade of change. But that doesn’t mean change is not happening.

According to Harvard political scientist Erica Chenoweth’s research, if you want to achieve systemic social change, you need to mobilize 3.5 percent of the population around your cause. Though we have not yet seen AI-related protests on that scale, we do have data indicating the potential for a broad base. A full 50 percent of Americans are more concerned than excited about the rise of AI in daily life, according to a recent survey from the Pew Research Center. And 73 percent support robust regulation of AI, according to the Future of Life Institute.

So, even though you might feel alone in your workplace, there are people out there who share your concerns. Find your teammates. Come up with a positive vision for the future of tech. Then, fight for the future you want.

Bonus: What I’m reading