Ideas That Change The World
They're self-explanatory, insightful, and polished.
When I taught Model UN, I'd always say delegates should strive to give speeches are the ones that change the course of debate. Not the longest, the most flashy, nor the best delivered. You need to get the basics right — you will not succeed if no one can hear it, or if people get bored halfway through.
But there is nothing more satisfying, nor impactful, then by dominating the conversation by only speaking once.
During committee, I'd try to achieve these through clever acronyms "Treaty on the Regulation of Uranium Mining and Processing" (before ChatGPT democratized access to such things), reducto-ad-absurdums ("this conference fits the house's definition of modern slavery!"), or other contrarian takes. But, in the global marketplace for ideas, that's not enough.
The best ideas light the discourse on fire.
How do you have an idea that changes the world? Some practical advice.
The Idea Must Explain Itself. The core concept, and your particular unique insight, should be summarizable in a single phrase. "The Intelligence Curse" (the resource curse applies to AI), "Situational Awareness" (you see the future first in San Francisco), "AI-2027" (AI, in 2027).1
Counter-example. "Shard Theory" is a terrifically compelling explanation of human learning and motivation that I have to re-explain every time. Also, don't put it on LessWrong.
Mimetically fit. Do I want to tell the ideas to my friends? And will they remember?
Descriptive, not prescriptive. The best essays show you something new about the world. They are written in that tone too: I know a secret, and I'm letting you in.
Polished. The project cannot be 80-20-ed. Draft, and draft again. Think about how well-written and beautiful both Situational Awareness and AI-2027 were.
Definitive Reference. What you produce needs to be the definitive, undisputed reference on the idea. Do the work.
Visual. Have pretty graphs that capture the core of your idea.
