What is AI and why and how is it relevant for the law? Thoughts from Daniel Martin Katz

On Thursday, May 18, IE Law School hosted ieLawX, an afternoon of discussions on the relationship between artificial intelligence (AI) and law. The opening talk was given by Daniel Martin Katz, technologist, lawyer, professor, and self-proclaimed “legal rebel.” He had 30 minutes to share his thoughts on what AI means and how it’s related to the legal field.

He started by giving us a brief recap of the history of AI in law. From the 1980s to about 2005, he told us, rules-based AI dominated. This includes tax preparation software like TurboTax; Neota logic; and the A2J author. From 2005 to now, data-driven AI has led the way. “We are now learning the rules and (…) trying to predict what someone will do by collecting massive amounts of data.” What classifies something as massive, you may ask? Katz referred to a petabyte, which is 1000 terabytes.

This brief summary was followed by Katz’s analysis of the current state of the legal tech industry. He explained that law is now trying to use data to predict elements of the legal process: costs, case outcomes, relevant documents, etc. Historically, humans have made predictions in three ways: by asking an expert, consulting a crowd, or building an algorithm. Which method has been the most accurate? Katz used the example of Supreme Court judge predictions: experts only get it right 58% of the time; crowds, 84%; and algorithms, 70%. So why not always trust the crowd? According to Katz, the “secret sauce” is to blend the three, because humans plus machines is much better than either one on its own.

Katz then moved to the issue of what the future holds. One of his assertions was that “the democratization of machine learning is underway.” Machine learning as a service (MLaaS) is growing in popularity, and we’re entering the golden age of the open-source model. So what are the implications of this shift? In Katz’s opinion, every organization—not just in the legal field, but across all industries—needs a data strategy and relevant human capital. It’ll be crucial to have these experts, even if it means growing that talent internally.

Katz summed up his thoughts with a simple and reassuring conviction: so far, there’s no AI technology that can do what lawyers do, and we are nowhere near reaching that limit.

The ieLawX event is a direct reflection of what IE Law School stands for: a cutting-edge approach to legal education and a steadfast commitment to innovation. The school promotes an entrepreneurial spirit in students so that they don’t just adapt to the rapidly changing world—rather, they’re the ones driving it.