Yao
Yiheng

"Science gathers knowledge faster than society gathers wisdom". Asimov wrote that decades ago, and it has only gotten more true. What keeps me up (in the near-term) isn't whether AI will be (too) capable — it's whether its benefits will reach everyone before the gap becomes irreversible. And then we can bask in the existential dread of superintelligence together...

Intelligence is fundamentally a search problem, and language is to minds what genes are to organisms — the medium through which minds replicate themselves. Experience is the ultimate ground truth, and some truths can only be understood by living them; Sartre, Heidegger, and Husserl's work embodied that. Furthermore, the scientific pluralism rooted in the pragmatist tradition reminds us that multiple, seemingly contradictory models of reality can simultaneously hold value if they prove useful in practice. We need more of Fitzgerald today, who recognized that the test of a first-rate intelligence is the ability to hold two opposed ideas in the mind at the same time, and still retain the ability to function.

When I'm not at a computer I'm usually building something with my hands, playing tennis, occasionally sailing a dinghy into the wind, or watching animated rectangles wage political war throughout history. I think about space-faring civilizations more than is probably healthy.

Yao Yiheng
Research Interests
  • AI Alignment & Interpretability
  • Human-AI Collaboration
  • Cognitive Science & Decision Theory
  • LLM Evaluation & Behavior
Background
Contact
  • https://github.com/devYaoYH
  • github username at google mail servers