Excellent @BryanGarnerLaw @ABAJournal essay on how AI is changing what lawyers do. Great read for law students and law faculty. ### If every lawyer can produce clean prose, clean prose will cease to impress. If every lawyer can generate a plausible first draft, the first draft is cheapened. And if every lawyer can produce a competent summary of the law, well, you get the idea. The real work will be judgment. And judgment doesn’t announce itself with good grammar. It appears in the choice of argument, the recognition of controlling authority, the treatment of inconvenient facts, the allocation of risk, the refusal to make an unnecessary concession, the decision to say less, and the ability to see what everyone else has missed. ### In law school, we need to move from teaching tasks AI is automating and focus on developing critical thinking and judgment while working with AI. From Day One, AI mentors and simulators can help law students learn more, faster and help them develop critical thinking and judgment. Law professors need to transform the way they teach and incorporate AI tutors, simulations, and other tools that they have customized and developed. Working with AI will be a necessary component of exercising judgment, and we underestimate the capabilities of AI to produce aspects of what we today call judgment. Nonetheless, there will be a need for a human to steer the AI and be accountable, and there are unlimited emerging problems and scenarios where humans need to be deeply engaged and accountable. Law students and faculty need to understand the difference between offloading work to AI, on the one hand, and iteratively co-creating with AI, on the other. Law schools on Day One need to introduce AI and teach law students to work with AI to be proficient in critical thinking and judgment at levels that are not possible without AI. The full @BryanGarnerLaw @ABAJournal essay should be required reading for all law students and faculty: abajournal.com/columns/artic…

"Compared to what" is a key question when we assess the impact of AI. We tend to romanticize the alternative. When assessing AI for legal tasks and judging, it is often assumed that humans are highly accurate and unbiased. Data tells us otherwise in many cases. Humans often fail the validity and reliability expectations that we apply to AI. We should interrogate both the impact of AI systems and the impact of the current human-centric ways of doing things. Regarding the use of AI for scholarship, we romanticize how humans research and write. As @Dschwarcz points out, many of the ways AI contributes to scholarship are very similar to how human RAs contribute, with few objections by academics. But can we say that the "generating of ideas" is "performed solely" by the human author? Even when AI is not used at all, research suggests that humans have a very poor understanding of the origin of their ideas. Also, when we use AI and human RAs for initial research and related tasks, it's likely that this affects the generation of ideas. "AI disclosure" serves no purpose and upon interrogation is counterproductive, IMO. It raises more questions than answers. It also creates a chilling effect for many. (Consider, who is permitted to disclose and who would be "punished" for disclosure even when the work is excellent?) To be clear, AI use should be permitted in scholarship and appropriate AI use should be encouraged. Used appropriately, AI improves both the efficiency and quality of scholarship, as @Dschwarcz said. For our students, we need to change the way we teach and assess to ensure that they use AI responsibly and learn. What interventions do we need in scholarship to address improper uses of AI? Perhaps we need more venues for oral and written engagement and constructive criticism of scholarly work, particularly in legal scholarship. In academia, if scholars are taking shortcuts to produce scholarly output, surely this will show up in due time. If scholars who use AI can defend their work and continue to produce high quality work in their field, then what is the problem?

Yes, I agree @OrinKerr, the answer to AI is changing the way we teach and assess. Moreover, AI can help students learn more, faster. For example, we can create AI tutors and simulations for students. AI amplifies capabilities and is raising the bar for everyone. Instruction needs to change, and there is a strong case for embracing AI across the board, especially in professional schools.