Why shaming people about AI slop isn’t enough to stop Big AI

These days, the conversation in tech and business, and in a lot of society, is still all AI, all the time. And one of the most fundamental questions boils down to: How do you get people to change what they’re doing in regard to AI? For people who (understandably) have moral or ethical objections to the many harms caused by Big AI, there’s the challenge of how to drive action while lacking the resources and capital of the tech tycoons who’ve driven the broad cultural push towards AI adoption.

As a result of the power differential between those pushing AI and those fighting its advances, the typical rhetorical tactic for AI critics has been to try to attach stigma to the use of AI, and to the outputs of AI systems. There is also little cultural discussion, or even mention, of alternative offerings that aren’t from the Big AI companies, so the entire narrative is framed as either using the most harmful, exploitative, damaging AI tools from the likes of OpenAI, or using nothing at all.

And for the most part, this has been pretty effective for people who have any taste or sense of culture. If you’re a creator, or engaged in creative culture, you probably either can’t stand the AI aesthetic, or feel betrayed when you find out something that was appealing to you was AI-generated or made by someone who used generative AI to create it. The only creative discipline that’s broadly an exception to this is coding, where for the most part people don’t have as many aesthetic objections, but even there, people pretty stridently object to the slop aesthetic in user interfaces and other human-facing aspects.

There are, of course, diehard cohorts of AI advocates who insist that they love the AI aesthetic and don’t mind the hyper-real look to what these systems output, but the mainstream discourse in the arts, design and creative disciplines reached a consensus some time ago, and it’s fairly stringently enforced amongst fan communities.

But shame isn’t an effective strategy for stopping the advance of Big AI as a force in society. I’ve said before that scolding people out of using these tools hasn’t worked and isn’t going to work. What I want to get at here is why it structurally can’t work — and what has to replace it.

Shame is already priced in

I wish shame and stigma were more effective ways of driving or changing behavior in society in general. It has been a project in many social circles over the last 15 years or so to remove shame as a social curb on bad behaviors, and this has largely succeeded; this is why we see the richest man in the world monetizing accounts that share child sexual abuse material on his social network, and the president of the United States having been found liable for sexual abuse, and no one holds these men to account, or is even unwilling to be seen with them. Having grown up fluent in an Asian culture, I’m well aware of how shame can shape entire communities’ functions and norms.

But in the particular case of how new technologies shape society, trying to use shame to drive behavior change isn’t an economically effective tool for curtailing the impact of bad actors. During the rise of the social networks, when we first started seeing the kinds of harms that they could cause, many activists and academics who wanted to champion alternative approaches tried to shame people out of using the then-dominant platforms. That stigma persists — you still see people today who are half-embarrassed when they talk about how they’re “still” on Facebook. But they are on Facebook.

People being embarrassed or ashamed while still using the technology does not bother the authoritarian owners of those platforms in the least, as long as they’re profiting off of those ashamed users.

That pattern repeated with the gig economy. Though the #DeleteUber movement early in the first Trump administration was briefly very effective at scaring Uber’s leadership, they quickly adapted to the criticisms people had, and most people have forgotten they were ever supposed to be angry in the first place. They knew they could weather the brief shame or stigma that was attached to using the app.

In short, the venture capitalists and CEOs of Silicon Valley have priced the costs of shame into their business plans. It may make creative communities feel good to shame people about using AI (and maybe that has some purposes for social cohesion amongst those groups), but it rarely stops people from using the platforms at all — it just makes them quieter about it.

Whenever I say this on social media, a bunch of really angry folks come into my mentions and say “you love AI and slop and why don’t you go give Sam Altman a hug” even though my bona fides of criticizing these tech tycoons go back to when the angry people were still in diapers. But I don’t fault people for having spare reservoirs of rage aimed at the harms that the authoritarian billionaires have done to their world; I just want them to aim their rage more effectively.

Are we being effective?

We have to care more about being effective than we care about being “right”. We have to care more about power and actually winning than we care about merely dunking on people. (And I mean, dunking on people is fun! It just can’t be the only thing. And it doesn’t win any converts.) I wish we would learn one goddamn lesson from our failures to stop the last several waves of tech adoption, where we’ve just kept repeating the same failing patterns, preaching to the choir with our platitudes while regular people all get fed into the insatiable user acquisition maw of big tech, social stigma be damned.

I’m lucky enough to be the kind of NYC policy nerd who saw Zohran Mamdani’s first campaign video when it came out, right after Donald Trump won a second term. It’s worth a re-watch, but in short, what he does is go to the Bronx and engage with folks in one of the areas of the city that swung hardest toward Trump during the election. And he just starts asking people what they care about, why they voted that way. One of the most remarkable things about the video is, Mamdani doesn’t argue with anybody, he just listens, and then he asks them if their life would be better if their rent were frozen, if buses were free, if childcare were affordable. Perhaps most relevant to this point, he doesn’t shame them, even though they had very obviously made a devastatingly self-defeating choice. He gave them an affirmative alternative to move toward, and in doing so, empowered both those people and himself to do something better.

This is the kind of solution that most AI critics are not allowing for right now when they only picture a shame-based corrective to individual people’s use of AI. It is possible to build human-centric, empowering tools and platforms that are owned by the people instead of by billionaire tycoons, that are responsible about the environment and with people’s data and privacy, that are accountable to the communities that they serve. If we engaged every person who made some AI slop and said, “would you rather have tools that help you make something amazing, and brought you into a community of creators?” I don’t think anybody would say no.

And truthfully, I think two big reasons why critics don’t use that framing instead are due to laziness and arrogance. It takes a lot more time and effort to engage with people over time and to bring them into community. It’s a lot easier to dunk on them, and then return to one’s existing community of peers who will praise you for saying how ugly that person’s slop was. But that won’t stop the Big AI companies.

We need a new theory of change if we care more about stopping the harms of Big AI than we care about feeling righteous. The good news is that it already exists, in pieces: there are tactics for winning the platform war, there are direct actions that ordinary people have already used successfully, and there are alternatives being built right now that are worth pointing people toward. What’s missing isn’t the alternative. It’s our willingness to offer it to people instead of just yelling at them.