Overview
CAN WE DESIGN MATTER THAT COMPUTES AND REMEMBERS? Artificial intelligence is advancing at breathtaking speed, but its appetite for electricity and computing infrastructure raises a fundamental question: Is there a more efficient way to process information? In this research proposal, Lev Neymotin, PhD, and Samuel Neymotin, PhD, explore a different approach: smart polymers engineered to perform analog computation while retaining memory within the material itself. Inspired by Professor Alexander Yarin's polymer research, the information-rich structure of DNA, and our earlier work on analog computing, we propose combining the natural dynamics of engineered materials with the precision and reliability of digital systems. The key idea is simple but potentially far-reaching: design matter to perform useful computation directly, rather than require digital computers to simulate every physical process. The proposal includes a practical Proof-of-Principle experiment to investigate whether such a material can compute, remember, and adapt. This is a research direction, not an accomplished breakthrough. But its implications for energy-efficient AI could be substantial. 🎬 WATCH THE VIDEO: https://youtu.be/L5Pfw2ZL2JM 📄 I particularly recommend reading the full technical proposal here. The video introduces the idea; the paper develops its scientific foundation, proposed originality, system architecture, and experimental approach. Lev Neymotin & Samuel Neymotin Uncompressed Thinking