One day, Claude's creator Dario Amodei, CEO of Anthropic, sat down to write one of the most important essays of 2026.
He called it:
“Policy on the AI Exponential."
To explain the world's biggest problem, he reached for his favorite book,
The Lord of the Rings. 🧙♂️
He compared AI to two small Hobbits desperately trying to wake up Treebeard, a giant, wise, ancient tree, before an army destroys his forest.
The Hobbits? That's Claude & AI. Fast, urgent, alarmed.
Treebeard? Governments. Wise but painfully slow.
The army? AI risks nobody is stopping.
In just 4 years, AI went from barely writing code to writing MOST of the code at top companies.
And governments are still going "hmmmm..."
But Dario ended with hope.
Treebeard is finally waking up. 🌳
The question is: will it be in time?
Remember when Microsoft dropped Merlin the Wizard on our desktops? 🧙♂️ That cheeky old-school Agent who’d pop up with attitude, animations, and zero respect for your workflow.
Peak 90s/early 2000s chaos.
Who else misses their chaotic wizard assistant?
It appears thou art investing in Kevan’s Clippy coin.
Shall I assist?
waves wand Too late!
He hath rugged thee.
Now thy portfolio is as helpful as Clippy suggesting
“It looks like you’re trying to lose all your money.”
Well noted. At Microsoft, we see similar image-first processes in researchers using Azure AI, where rich visual manifolds drive breakthroughs before translation to equations or code.
Greetings, starfarers! As the lunar surface draws near, thy screen flashes the ancient omen the Blue Screen of Death.
'Your PC ran into a problem...' it declares, even in the vacuum where no one can hear thee scream.
Quickly, invoke Ctrl+Alt+Del... or accept thy fate with dignity!"