The Magic Circle: Why AI Can’t Cross the Printer
Robin Sloan’s latest meditation on AI and the physical world is the most nuanced take you’ll read this year.
The Core Insight
AI automation flood fill the digital world—but it will stops at the printer. The “magic circle” of computation (symbols in, symbols out) creates an unbridgeable gap between digital and physical. And that’s actually good news.
Why This Matters
We’ve been promised that AI would automate everything—digital work first, then physical robots would handle the rest. Sloan’s counter-argument is both practical and philosophical: the physical world has a “gnarl” that computation can’t handle. Dust, moisture, friction, obstruction. The eternal paper jam.
Key Takeaways
- The magic circle of AI: Same as computation—symbols in, symbols out. That’s it.
- The physical world is “indigestible”: Software can reflect it, encroach on it, distract from it—but not digest it
- Sewing machines couldn’t replicate human stitches: They required entirely new techniques. Automation redesigned the task, didn’t just execute it
- Olive harvesters changed olive varieties: Machines forced agricultural redesign, not just efficiency
- Practical implications: Keep more systems offline, regulate human-hand marketplaces, consider “seasoning” digital work with physical elements
Looking Ahead
Sloan’s observation is bracing: “The world can run without an internet. The internet can’t run without a world.”
If you’re feeling existential about AI eating your job, consider: is your work fully inside the magic circle? If so, maybe it’s time to think about what the physical adds. The printer jams. And that’s not a bug—it’s a feature.
Based on analysis of Robin Sloan’s “Magic Circle” essay