Demanding Curiosity

LLMs. So what?

If you think we should embrace these tools: How? Where? With what guardrails?

If you think we should resist: How? Where? What are you protecting?

If you think the answer is nuanced—good here, bad there—how do we tell the difference?

These questions are hard to answer.

1 of 4

Footnotes

"Legitimate" here in the sense of mutually justified. The term draws from legitimate peripheral participation (Lave & Wenger)—the idea that newcomers learn by entering practices at the edges, surrounded by people who've been at it longer. We pursue LPP in our Curiosity Builds sessions and borrow the language here to highlight a gap: there's not yet broadly shared respect for active, humble curiosity about AI across pre-established positions.
↩︎
Virtuous curiosity, if you like—the kind that stays accountable, shares what it finds, invites challenge, recognizes boundaries. Not idle, meddlesome, or otherwise vicious. Curiosity itself must be moderated.
↩︎
The handoff concept comes from Mulligan & Nissenbaum's work on what changes when functions transfer from humans to machines—functional equivalence doesn't mean ethical equivalence. See the Research Grounding & References in our AI Principles for more on this framework.
↩︎