Science Trust Project Community Hour on Large Language Models and Wikipedia

Quality is a process – it’s difficult to verify the quality of information without examining the process by which it was constructed. This concept is familiar in scientific practice (see “methods”), but it might be less intuitive how it applies to Wikipedia and applied large language models. In this talk, Dr. Halfaker will bring insights from decades of working with Wikipedians to build a sustainable, high-quality information resource, as well as a couple of years working with the most advanced AIs available today (GPT).