At Vanderbilt Law, the team at the Vanderbilt AI Law Lab (VAILL) has a Substack titled “The AI of Law” where they document and reflect on the community’s efforts to learn about and teach about intersections among legal education, lawyering, and justice in the AI era. (The main VAILL site is here.)

From a recent post, titled “Plant Seeds, Not Gardens: What One Hour of AI Education Can (and Can’t) Do,” reflecting on VAILL’s first 1L AI Bootcamp, an elective one-hour mid-semester introduction to AI for first year law students:

We landed … on a one-hour bootcamp, strongly encouraged but not mandatory, scheduled later in the semester during the same week 1Ls would use AI tools in their legal research courses. That structure worked for our students, but it also brought up something we’d seen repeatedly in our work with practitioners: even when everyone agrees the training is essential, finding the right moment for it is its own challenge. …

From our upper-division courses, we know that students develop sophisticated AI judgment through experimentation: they need to try tools, see what works and doesn’t, compare outputs, and iterate their approach. They needed time to fail safely, to be surprised by both capabilities and limitations, and to develop their own evaluation frameworks.

The bootcamp format gave students a map, but not the journey. They saw demonstrations but didn’t get to drive. And that matters, especially with a technology where meaningful competence comes from use, not observation.

This isn’t a shortcoming of the content or the participants at all. It simply reflects the natural limit of what a one-hour format can accomplish when the goal is deep skill development.