Every two weeks I meet new scholars and fellows while reviewing the VALIANT applications—virtually at first, and later in person—who are eager to embark on their own translational AI journey. Our community is full of brilliant, creative people, and we’re fortunate to keep welcoming new scholars and fellows into the fold. From the transformative NIH K awards to more than twenty peer-reviewed articles in print this month, the pace of high-profile innovation continues to amaze. Our scholars arrive full of questions, one of the most common being: “What do I need to do to get a paper?” 

It’s a fair question, and one that speaks to ambition. But I think we sometimes misunderstand what a paper really is. A paper is not a victory lap or a trophy. It is a gift—knowledge distilled and shared with others. Every setback on the way there offers us the chance to refine our hypotheses, sharpen our tools, and deepen our understanding. If every algorithm worked perfectly the first time everytime, we wouldn’t be doing science. In translational AI—a young field where approaches are still being tested and boundaries still being drawn—the real advance comes not just in what works, but in why some ideas work, and why some don’t.

I often remind myself of three great types of papers. The first proposes a new theoretical lens or optimality criterion. The second creates an empirical system that does something no one has done before—or does it substantially better. And the third, often overlooked but equally important, explains why a method is effective and brings clarity to the state of the art. One of my doctoral mentors even celebrated this spirit through the “Hairball Award”—given not to the most cited paper, but to the bold idea that branched off from the wavefront of knowledge and opened a fresh way of thinking.

So let’s remember: the true currency of our field isn’t the PDF itself, but the teachable moment it represents. That might take the form of a journal article or conference paper, but it might also be a GitHub repository, a vibrant scientific discussion online, or a workshop conversation that sparks new insight. Each is a way of teaching, learning, and carrying knowledge forward.

This summer, I’ve seen that spirit alive in our AI Scholars—learning from each other, from fellows, from guest lecturers, and from faculty alike. We’ll carry that momentum into my upcoming Studio in AI course this spring, where formal lectures meet hands-on mentorship and peer learning in dynamic spaces like FGH 203 and our high tech GPU-teaching environments.

As we look ahead to our academic retreats and the conversations they will bring, I’m grateful for the opportunity to continue building connections across Nashville and beyond. These are times to break the mold, to think differently, and in so doing, to reshape how we approach AI and society. I look forward to the road ahead, and to hearing from—and learning with—all of you.

Best wishes,