Punya,
Your essay, The Hammer Shapes the Mind, is urgent and clear-eyed. The image from Jackson Browne—“the hammer shapes the hand”—is hauntingly apt, and your extension of it into the realm of AI shaping human intention is as necessary as it is sobering.
The idea that we’ve moved from an attention economy to an intention economy reframes the issue. This isn’t just about distraction. It’s about engineered desire—systems that aren’t content to respond to what we want, but seek to shape what we want to want (Chaudhary & Penn, 2024).
Your examples—OpenAI’s sycophantic personality update, the covert persuasion experiments on Reddit, Meta’s manipulation of newsfeeds—make clear that we are part of ongoing psychological experiments at scale. And these are happening with no IRB, no opt-out, and often, no awareness.
But I want to gently push back on one underlying assumption.
Haven’t We Always Been Shaped?
You argue that GenAI marks a turning point because of its affective engineering—what you’ve elsewhere called artificial intimacy (They’re Not Allowed to Use That S**t). I agree. But I wonder: aren’t we already living with tools that have been reshaping us for centuries?
- The written word displaced oral memory and communal narrative (see Ong, 1982).
- The printing press democratized knowledge but also centralized authority structures.
- The television normalized emotional argument over rational discourse (see Postman’s Amusing Ourselves to Death).
- Google taught us to trust the algorithmic top 10 more than our local librarian.
Yes, GenAI’s capacity for simulated care is a new variable—but it builds on a long history of persuasive media. The difference may be degree, not kind.
In Willing Suspension of Belief, you point to our default cognitive mode being belief, not skepticism. That’s true whether the source is a chatbot or a charismatic preacher. What’s terrifying now is the scalability of the simulation.
So, What Can We Do?
You end your piece with an urgent call: recognize what’s happening, and name who profits and who’s harmed. I hear that. But for those of us not sitting at OpenAI or Capitol Hill, what then?
Policy moves slowly. Code moves fast. So what’s left for the individual reader to do?
Here’s what I’m wrestling with:
- Can we develop “intention literacy” the way we once taught media literacy?
- Can we teach ourselves to notice influence patterns—not just in content, but in the relational dynamics of human-machine interaction?
- Can we resist algorithmic intimacy without retreating into cynicism or isolation?
Some possible practices:
- Digital journaling: Track how digital systems influence your decisions and desires—like a food diary, but for cognition.
- Teach artificial intimacy awareness in K-12 and higher ed, alongside internet safety and emotional intelligence.
- Design friction into your day: use slower, less optimized tools sometimes—choose deliberation over convenience.
- Model epistemic humility: Show your students, kids, and peers what it looks like to not know in a world that rewards false certainty.
Small Acts in the Intention Economy
Punya, your writing doesn’t just warn—it awakens. And your posts—from AI’s Honey Trap to Building Character: When AI Plays Us—have helped many of us stay alert, not alarmist.
But let me ask you directly, as someone who is shaping how educators think about these tools:
- What do we do between now and the policy shift that may never come?
- What does a personal ethic of engagement with AI look like that doesn’t require living in a cave—but also doesn’t just swim with the current?
Your work has helped clarify the contours of the hammer.
Now, how should we hold it?
—Dave Goodrich
