Why Does Scaling Work So Well?
More data, larger models, more sampling, more search—and capability keeps emerging. Understanding the recipe has only made it stranger to me.
Essay on scaling and generative intelligence Sep 2026
Welcome to the World of Intention
If AI can do almost anything, then the interesting question is: what do you want to happen?
Essay on intention Aug 2026
How Science Compresses Complex Systems
Thinking on how science evolves with its minimal description length: scientific progress as the search for what can be forgotten without losing the phenomenon.
Essay on scientific compression Aug 2026
Be Unbenchmarkable in the Agent Era
Why benchmarkable skills are less valued, and why being unbenchmarkable is more necessary in the AI Agent Era.
Essay on human agency May 2026
The Unreasonable Elegance of AIGC: On the Learnability of LLMs and Diffusion Models
How next-token prediction and score matching turn intractable generation into tractable, atomic tasks—and why the resulting optimization landscape makes modern AIGC both possible and elegant.
Essay on learnability & AIGC Feb 2026
On the Surprising Coincidence Between Large Language Models and Ordinary Language Philosophy
A philosophical essay on meaning as use, normativity, and judgment—and on what contemporary large language models reveal about the power and limits of linguistic fluency.
Philosophical essay Dec 2025
AI Theory: Fragmented but Expansive
I predict AI won't converge to a single unifying theory, but rather to a rich tapestry of frameworks—each illuminating different facets of learning, reasoning, and intelligence.
Personal prediction on AI theory Aug 2025