A few months ago, absolutely everyone was hearing and talking about the study from MIT titled “Your Brain on ChatGPT”. It’s a brilliant bit of branding for virality, but I have some key problems with how the experiment was designed - the impact of which could’ve made the public conversation that followed something actually productive for society.

When I first read the study, my gut reaction was that it wasn’t controlling for how people actually use these tools.

Take essay writing as an example. Even with Google, you could just find an article, copy and paste chunks, and call it a day — that’s bound to lower brain activity. But you could also use ChatGPT as a thinking partner: bouncing ideas back and forth, challenging your assumptions, refining your argument. That’s a completely different process, and one that would likely show much higher brain activity.

My view on LLMs has always been that the magic, and the danger, comes from how we use them. They’re not at their best when you’re chasing the end product. But if you use them as collaborative tools, to assist your creative journey, they can expand the way we create by bringing us alternative perspectives and helping us critique our ideas. This study didn’t touch that usage nuance at all.

The study is going viral. Really viral. And the headline message it leaves people with is: “We’re doomed.” But that’s not good enough of a takeaway.

In Capitalist Realism, Fisher writes that today’s media and entertainment enable us to perform worry about our world’s problems. We’re left satisfied that we are indeed worrying about the problem, without needing to take any action.

By missing the nuance about how we use AI tools, the conversations on this study remained surface level. “This is why I don’t use it!”, “I don’t care, I’m still going to use it”, “bye brain”.

Instead, the study could’ve sparked something useful by asking: “When do these tools sharpen our thinking, and when do they dull it?” That’s the kind of insight teachers, teams, and policymakers can actually act on.

AI regulation isn’t going to appear out of nowhere, but it’ll show up if enough of us have a more nuanced understanding the pros and cons and demand thoughtful guardrails.

If we want AI that serves people rather than just profit or panic, we have to practice and promote nuance. Let’s talk about how we use these tools with our friends and coworkers. Ask your school or team to publish an “AI use” policy that prefers partner mode for learning and decision-making. Tell your representatives you want transparency, labeling, and audit trails baked into law.

Curiosity over fear, design over vibes, and collective pressure over performative worry - that’s how we get the tech, and the rules, we actually deserve.