Interview with Akari Asai – beyond scaling: frontiers of retrieval-augmented language models

Summary

In this crosspost from AI Matters, a publication of ACM SIGAI, Ella Scallan interviews Akari Asai, winner of the 2025 AAAI Doctoral Dissertation Award. Asai's doctoral research focused on retrieval-augmented language models (LMs), arguing for moving beyond monolithic LMs to augmented LMs that are designed, trained, and deployed with complementary modules. She pioneered retrieval-augmented LMs, which retrieve relevant knowledge from large-scale text collections at inference time rather than relying solely on memorized training data, enabling more accurate, up-to-date, and verifiable outputs. Her work challenged the prevailing assumption that simply scaling models would solve hallucination and outdated knowledge issues, offering a more effective and efficient alternative. This approach led to OpenScholar, an open-source model aiding scientists in synthesizing vast literature, and has been widely adopted in academia and industry, powering systems like generative search engines.