ChatGPT is our NEW DOPAMINE

On June 4, 2024, I was two cans of Red Bull deep, with only two hours left to submit my coding project. As usual, I was locked in a familiar routine with my best friend, ChatGPT: ask it to write code, test it, ask it to fix the bugs, and repeat.

Then, suddenly, the prompt froze. ChatGPT went down.

It was a global server outage, and I was completely helpless. With time running out and no AI to save me, full panic set in. My back started sweating, my hands shook, and my heart raced. In that moment of pure anxiety, a cold realization hit me: I was going through withdrawal, I was addicted to ChatGPT.

The physical symptoms were no different from those of a drug addict losing their supply. It became clear to me that I wasn’t just adopting AI as a helpful tool anymore—not only me but many of us are slipping into a dependency on the frictionless, instant dopamine it delivers.

Our Fried Dopamine Circuit on AI

Evolutionarily, our dopamine system was designed to reward hard work. The friction of hunting for food or solving a complex problem made the final reward meaningful. But technology has rapidly removed that friction. Stanford professor and psychiatrist Anna Lembke notes in her book Dopamine Nation that modern society provides endless tools that deliver unearned dopamine hits—giving us the reward without the effort. Today, generative AI risks becoming our ultimate intellectual shortcut. We no longer wrestle for hours trying to articulate a complex thesis statement. We simply type a prompt and receive a polished output instantly, getting hooked on the frictionless gratification of a final product while bypassing the intellectual struggle altogether.

This addiction to instant gratification is fundamentally changing the value of human knowledge. OpenAI CEO Sam Altman recently noted that we are heading toward a future where "intelligence is too cheap to meter." When generating a flawless output costs almost nothing, the incentive to learn through struggle disappears.

For me, as an international student, the opportunity cost of mastering academic English suddenly feels overwhelmingly high. Why spend hours agonizing over vocabulary, sentence flow, and nuances when an AI can polish my thoughts into a perfect, native-sounding essay in seconds? It feels like driving on a fast highway where everyone else is speeding ahead with AI, and the Fear Of Missing Out forces us to keep hitting the gas just to keep up. But in our rush to stay competitive, we are quietly outsourcing our brains.

Bringing Our Own Intelligence to the Table

In my studies of Human-Computer Interaction, the ultimate goal isn't just building smarter machines; it’s figuring out how to work alongside them. Wharton professor Ethan Mollick calls this "Co-Intelligence"—the idea that AI should amplify human thought, not replace it. But to truly collaborate with AI, you must first bring your own intelligence to the table. This reality hit me hard when I transitioned from the tech-heavy environment of KAIST to Wellesley College.

The Human Constant

At Wellesley, mentioning “AI” often sparks immediate skepticism. On the surface, this might look like a general distaste for big tech. But beneath Wellesley’s fiercely analog culture lies a profound awareness: the understanding that bypassing the struggle of learning diminishes our intellect.

In our "device-free" classrooms, students print out readings and engage in rigorous, face-to-face debates. Coming from a tech background, this initially felt like a stubborn resistance to efficiency. Yet, it turned out to be exactly what I needed. By wrestling with paper materials and defending my arguments without an algorithm to save me, I relearned how to think independently. Paradoxically, embracing this analog constraint is our greatest advantage. When AI makes everyone artificially smart, the only variable that matters is the human constant.

Weighted Vest

We cannot stop the AI highway, but we can choose not to let our mental muscles waste away. In an era where AI excels at almost everything, choosing to work without AI is incredibly difficult because we are so vulnerable to its addictive convenience.

Yet, embracing this intentional constraint is like training with a weighted vest - the resistance builds the strength needed to master the tool later. In the future, the people who truly excel will not be the ones who merely relied on AI to generate easy answers. They’ll be the ones who have trained themselves to think independently, critically, and creatively—even when the screen freezes and the machine goes dark.

American Studies 355
Calderwood Seminar in Public Writing: Critiquing American Popular Culture
Paul Fisher
Wellesley College, Spring 2026