Artificial intelligence (AI) and quantum information (QI) science are among the most active areas in cutting-edge science and technology, addressing the computational complexity frontier. Although these two domains have evolved separately in the past, recent breakthroughs in both fields create a special opportunity to tackle outstanding challenges through AI-Quantum interdisciplinary research. QI science is entering a new era with the recent realization of error-corrected logical qubits and logical quantum processors, enabling quantum algorithms of unprecedented complexity. To deploy these systems as new tools for scientific discovery, there is a burgeoning need for sophisticated methodologies to encode and decipher quantum states, efficiently correct errors, and seek implementations that can benefit specific scientific applications. 

In this talk, we will explore using mechanisms of large language models to address outstanding challenges in quantum information, such as state characterization and decoding. The key guiding insight is to focus on the space-time geometry of the measurement data for a given problem. This data-centric perspective offers a new angle to the fundamental problem of measurements while solving practical bottleneck issues of certification and decoding of states.