There is no accepted view about how technology should be taught in law schools. Our view is that this is a good thing. Law schools presently have no obvious ways to distinguish themselves in the market for students. There is no right way to approach this new world and the variety of responses to AI will give law students a range of options. At Duquesne Kline, we have very consciously chosen to create electives that educate students in the foundations of computer science and statistics. The classes are expressly designed for those with no background in these areas but require very intense work as we teach these courses very much as computer science courses with legal applications. It is our belief that if students learn fundamental math and computational thinking, they will quickly adapt to a range of software designed for lawyers. In our view, they will also have the background knowledge to interface with experts in the fields of computer and data science, something that will be essential in a world in which business and litigation are driven by big data. Finally, students who have worked on writing code from scratch to address legal issues will have the insights to think about how legal technologies can be designed to improve the work of lawyers and increase access to justice.

We propose to discuss the two of the doctrinal classes we have recently added to our curriculum and how the skills from those courses have been integrated into our clinical program. Wes Oliver, who teaches the two doctrinal courses, will talk about the virtues of imparting this knowledge to future lawyers in the abstract. This part will constitute 30 minutes of the program. Kate Norton, who teaches in the Duquesne Kline Family Law Clinic, will talk about how students with this knowledge have created a computer program to assist pro litigants as well as student attorney dealing with custody appeals.

The doctrinal courses:

In 2022, we introduced Coding for Lawyers, a course in Python programming in which students learn the skills one would typically acquire in an undergrad course on the subject but with projects aimed at improving the legal profession. We have been experimenting in the last two years of the course with identifying optimal assignments that contribute to the needs of the legal profession and stretch the students computational skills. At present, the final projects involves writing a code that extracts various types of terms from multiple contracts and collects them in one spreadsheet, requiring students to develop a system that can identify ways various contractual term can be identified and implement a system to search for them. This session would demonstrate the skills learned and how they were implemented to create the final project.

In 2024, we then introduced a course called Statistics and Machine Learning for Lawyers. This course deals with many of the topics that would be covered in an undergraduate course on the R language. The Coding course emphasizes thinking about the way computers process information while this course emphasizes what computers work with — big data. Concerns are frequently raised about implicit biases being replicated by AI devices, but all too often those concerns are raised in an alarmist manner that suggests such devices should simply be banned. A sophisticated understanding how how computers work with big data to make predictions provides students with the ability to raise concerns specifically identify how biases enter into the prediction, as well as the virtues of such predictive algorithms, ultimately with an understanding of how to balance the risks and possibilities of a given predictive tool. This part of the program would look at the predictive policing model students designed through the course of the semester as well as the very specific criticisms students were able to identify in the model after they had come to thoroughly understand the data, the statistics, and the computer programming that went into its design.

The Clinical Implementation

During the 2022-23 academic year, as we were introducing the first Coding course, our family law clinic received a grant to develop a piece of software to assist pro se litigants in appealing custody decisions. Working with students in the clinic, our students designed a device that would allow pro se litigants to determine whether a draft of an appeal satisfied the procedural and jurisdictional requirements for an appeal. Students in the clinic used the tool to determine whether appeals that had been filed were procedurally defaulted, allowing the clinic to preserve its resources in representing clients with viable claims. The device also identified other avenues of relief, permitted the student-attorneys in the clinic to quickly triage cases. As most of the students in this clinic had taken the Coding for Lawyers class, they were able to be a part of the computational design of the code as well as troubleshooting and modification as the project continued.

This session will discuss how computational thinking is not only valuable in itself, but more broadly how design thinking, required for programming, is very useful for lawyers as we have observed in the problem solving skills of clinical students. Many schools that have begun to teach technology to law students have chosen not to offer programming courses where students write code from scratch, but do very teach design thinking with software purpose built for legal task. The virtues of design thinking for lawyers that we discovered all very much apply to these courses.