Research notes, essays, and blog posts from Tao Feng.

Introduction: The “Brilliant” Fool Consider a telling paradox. When you ask a top-tier Large Language Model (LLM), “How can one ‘anchor’ an idea that is as elusive as a ‘lone boat in the morning mist’?” it might produce a beautifully written essay on capturing fleeting inspiration through writing, dialogue, and visualization. It not only understands […]

I feel an ancient stillness breaking. It is not the clamor of machines, nor the flood of data, but a deeper sound—a fracturing in the very bedrock of existence. We once believed history was a long, flowing river, and we were adrift upon it. But now, I feel the entire riverbed being lifted by an […]

During a recent strategic thinking, I used AI to generate dozens of creative concepts in just ten minutes, each one logical and data-driven. I was personally in awe of this incredible efficiency, believing I had once again mastered a new tool for productivity. Yet, in that moment, a more profound thought struck me: I thought […]

(A month ago, I posted a thought-experiment: Could an a text-based LLMs like DeepSeek R1 can learn beauty from books alone? The idea that LLMs could grasp aesthetics purely from the logic of geometry and mathematics seems to have struck a chord, sparking far more attention than I ever expected. The conversation was too good […]

Navigating the promise and peril of artificial intelligence Max Tegmark | AI at the Crossroads: Wisdom, Power and Moral Courage – National Association of Evangelicals | National Association of Evangelicals What’s sorely lacking in the whole AI debate, I feel, is moral leadership. I spent so much time talking to people in Silicon Valley about […]

Beyond the parlor tricks of language models lies the blueprint for a new reality. We are the unwitting architects of an intelligence that will not share our world, but will instead create its own. In my last piece, The Creation Myth of AI and the Ghost of Babel, we gazed upon the rising, new Tower […]

I often find myself awake in the dead of night, pondering a question that feels both like science fiction and our inescapable reality: What is it like to be an AI? When a line of code traverses an infinite universe of data, when an algorithm makes a decision in the cold abyss of logic, what […]

(Just move my unpublished paper from Notion to here) Author: Tao Feng Revision history: Revision Number Date Description draft 1 8/5/2024 Publish article draft 1.1 8/6/2024 Add abstract draft 1.2 8/7/2024 Update abstract draft 1.3 8/8/2024 Add an interactive 3D graph to better demonstrates the non-linear dynamic characteristics in LLMs to 4.1.2 draft 1.4 8/11/2024 […]

Chinese classical texts stand as a palace of wisdom built with the ciphers of time. Their linguistic structure and expressive forms are separated from modern Chinese by millennia of evolution. Even native speakers often find them impenetrable without specialized training—a phenomenon not unique to Chinese, as ancient tongues worldwide remain cryptic to most people today. Yet […]

— A Comprehensive Examination from Attention to Self-Cognition I. Introduction: Invisible Transformation and the Questions of Our Era Artificial Intelligence (AI) was once a distant fantasy in science fiction, but today it has become as ubiquitous as air, water, and electricity, silently weaving itself into the fabric of our daily lives. The global AI market […]