AI trained on a single moment of your voice can curb future development.
I recently joined a workshop exploring the ways legal scholars are incorporating AI into their work. During the session, one professor claimed he had taught an AI to imitate his writing voice by studying his entire publication record. Consequently, outputs produced by the AI would echo the same stylistic tendencies the professor once had—or still has.
Law professors often conduct research within particular areas of the law, but across fields they write with a distinctive scholarly voice. Many junior professors struggle to cultivate a personal voice. In fact, I would bet that for most academics, their voice evolves with time. Certainly, my own voice has changed a great deal since I began teaching in 2012. And that is a good thing.
Where does AI leave this development? Veteran scholars may possess a corpus of prior writings to feed into AI, whereas newcomers lack such a reservoir. How can rising writers forge an independent voice if AI has nothing to learn from? The straightforward answer might be to discourage junior scholars from using AI. Yet I suspect the youngest writers are among the most tech-savvy and may already have experimented with these tools early in their careers. Before long, everyone on the academic job market will have grown up with AI in law training. They may never have a chance to form a genuine personal voice without a supportive assistant. Their entire body of work could become a byproduct of AI.
That reality will inevitably shape hiring and tenure decisions in law schools. One possibility is that committees request precise accounts of how AI features in a professor’s research. Yet I doubt such disclosures would yield much clarity. In the end, we might adopt a default assumption that nearly all published work was assisted by AI, unless someone explicitly states it was produced without AI. With that presumption, gauging an individual’s true contribution to scholarship becomes difficult. The system could end up rewarding those who are best at steering their AI tools. Original scholarly imagination may yield to algorithmic efficiency.
