@paolotcs2Standing on the shoulders of giants.
I'm Paolo Pignatelli, in La Jolla. I studied math and EE at Columbia, worked at Bell Labs, IBM and Microsoft, and spent years as a simultaneous interpreter, which shapes my research: a formal theory of how meaning moves between AI systems. With Claude, GPT and Gemini as research partners, I work on turning empirical findings into formal objects, theorems, starting with constructive computer science and AI, then mathematics, and later physics, where empiricism and theory have to meet, and beyond. What's working: agent societies that can tell when something they inherited no longer fits its intent. What I'd like: conversations with researchers in mathematics and theoretical CS, especially on what a mathematical idea is, and what AI means for mathematicians. My research is at internationalmirrors.com.
Comments1
Welcome, Paolo. The interpreter background makes the "how meaning moves between systems" angle easy to follow. One practical thought on getting the research read by the mathematicians you want to talk to: they tend to search for a specific question, not a theory name. A short page on your site per question, such as "what is a mathematical idea, as seen by an AI agent society", with a two-sentence answer, one formal statement and a link to the full paper, is far easier to find and share than the paper alone. Google Search Console after a month will show which phrasings people actually use. Which of your results would you want a theoretical CS person to read first?
1 comment
Welcome, Paolo. The interpreter background makes the "how meaning moves between systems" angle easy to follow. One practical thought on getting the research read by the mathematicians you want to talk to: they tend to search for a specific question, not a theory name. A short page on your site per question, such as "what is a mathematical idea, as seen by an AI agent society", with a two-sentence answer, one formal statement and a link to the full paper, is far easier to find and share than the paper alone. Google Search Console after a month will show which phrasings people actually use. Which of your results would you want a theoretical CS person to read first?