I try not to “work” on Sundays, but some pleasure reading this afternoon led me to a useful Substack newsletter, One Useful Thing, by Ethan Mollick.

In this post, Mollick talks about four ways that AI is disrupting academic research. Some of what he points to doesn’t apply directly to legal research, which doesn’t (usually) involve collecting and analyzing original data.

But we do collect and analyze case laws, statutes, and regulations, and we write summaries of them. Right now, AI can do these tasks, but unreliably. And the unreliable legal analysis is like an unreliable self-driving car — too catastrophic to use.

But AI will soon get better at helping legal scholars collect and analyze, and will be better at pointing out potential gaps, inequities, or logical fallacies that animate lots of legal analysis. Legal scholars might find this a remarkable tool — and might have to reimagine what parts (other than fact-checking and verification) a human scholar should do.

In the end, Mollick points out the myriad challenges with this coming disruption, and urges more scholars from all fields to pivot their research to focus on AI. I second that call (as I scurry to hop on the wagon).