I don’t know about you, but I am sucker for technology and “Big Data” stories. I was glued in front of my television when IBM’s Watson took on Jeopardy’s reigning champion Ken Jennings–and won. I am interested in the work of scholars such as Dan Katz and initiatives such as ReInvent Law™ and LawWithoutWalls.™ When NPR and the New York Times ran stories about how technology may do a better job than lawyers for certain tasks such as e-discovery, I emailed those stories to friends and colleagues. My ears perk up when I read about the coming “disruption” to the legal profession. I often recommend to others Richard Susskind’s book entitled The End of Lawyers? about the impact of technology on legal services.
Regardless of whether you share my fascination for these kinds of topics, I encourage you to read the article entitled The Great Disruption: How Machine Intelligence Will Transform the Role of Lawyers in the Delivery of Legal Services which Professors John McGinnis and Russ Pearce contributed to the Fordham Colloquium on The Legal Profession’s Monopoly on the Practice of Law. In my view, regardless of one’s field of expertise, everyone should read this article and begin to reflect on this phenomenon that will revolutionize the practice of law, and dramatically change all of our lives.
This article begins by describing the rise of machine intelligence. It provides details that may be as new to other readers as they were to me. For example, IBM has created a division around Watson, has invested $1 billion in its development, and is using its program to aid in medical diagnosis. The authors explain that in a “recent competition on how to make use of Watson, the winning entry centered on the legal field, using Watson to search for relevant evidence in data and predict how helpful the evidence will be to winning the case.”
After noting acceleration in the development of machine intelligence, the article identifies five areas of legal practice that will change dramatically in the near future as a result of machine intelligence: (1) discovery; (2) legal search; (3) document generation; (4) brief and memoranda generation; and (5) prediction of case outcomes. The section on discovery, for example, focuses on the role of predictive coding in large scale litigation, which has already been used in some antitrust cases with the U.S. Department of Justice’s approval.
The section on legal search explains not only that machines will soon have the capability to search all legal precedents, but also notes the expanding role of semantic search, the increased ability of machines to make “judgments” about the strength of precedents, and their future ability to identify the issues implicated by a given set of facts. It predicts that machine intelligence will replace the legal search function of lawyers.
Concerning document generation, the article suggests that as machine intelligence grows, computer-generated forms will have an even broader scope in legal practice. The authors predict that within ten to fifteen years, computer-based services will routinely generate the first draft of most transactional documents. The authors cite the example of a law firm based in Silicon Valley that has developed a program that automatically creates the documents for incorporating startups. This program has significant reduced costs by reducing the billable time from an average of 20-40 hours to a handful of hours: “In cases with even extensive documents, we can cut the time of document creation from days and weeks to hours.”
The article predicts that although briefs and memorandum are more difficult than forms, in the future, machines will become more and more adept at generating useful first drafts. This section discusses the current ability of machine intelligence to generate news stories based on data and the likelihood that there will be continued improvement in this area.
This fifth development the article discusses is predictive analytics, which is a new discipline that combines data with analysis to make predictions. The article notes that using big data to guide decisions is one of the most important trends of the last decade and that it has now come to the field of law. Professor Pearce and McGinnis’ article provides several examples of the use of legal analytics. These examples include Lex Machina, which has gathered data from thousands of patent litigation cases and is being used to predict outcomes in that field. They also cite the 2002 Supreme Court Forecasting Project, in which a model of Supreme Court decision-making, which was based on previous Court decisions, more correctly predicted outcomes than a set of Supreme Court experts.
The concluding sections of the first part of the article are entitled “Future Trends” and “Answers to Common Objections.” The authors predict that because machine intelligence will permit clients to better compare price and performance, lawyers may lose market power to non-lawyers who provide what traditionally might have been thought of as “legal services.” The authors note, however, that “super-stars” may earn even greater returns than previously because data will make it easier to determine who is a superstar and because these superstar lawyers may use technology to deliver innovative solutions to problems faster and to a broader range of clients. The “answers” section anticipates criticisms about the predictive value of big data and about the degree to which machine intelligence will be able to replace human judgment. Anticipating those who might argue that the legal profession has always adapted to technological change (e.g., from carbon paper→copiers or from pens→typewriters→computers), the authors respond that machine intelligence is a different kind of technological change than previous technological changes because it substitutes technology for core legal skills and because the rate of change is faster than previously.
Having addressed the nature of the changes in the first part of the article, the second part asks whether regulatory barriers are sufficient to prevent machine intelligence from undermining lawyers’ traditional monopoly. It concludes that these regulatory barriers will not hold back the advance of machine intelligence. The authors find significant the degree to which U.S. consumers and businesses already are “voting with their feet” by selecting individuals or services other than lawyers to provide what traditionally was regarded as legal services. They also observe that the legal service market has become de facto deregulated with regard to machine intelligence and that increased enforcement of unauthorized practice of law is unlikely to stem the delivery of legal services through machine intelligence. Among other things, the authors point out the degree to which the legal profession already has accepted machine intelligence as an “input” and changed the regulatory rules to accommodate this input, citing the recent amendments to the comments of ABA Model Rule of Professional Conduct 1.1 – Competence. The article disagrees with commentators such as Larry Ribstein, Gillian Hadfield, Ray Campbell and Bill Henderson to the extent those commentators have argued that unauthorized practice of law or UPL provisions will hinder the application of machine intelligence to legal services.
The third and final section of the article discusses why opposition to machine intelligence delivery persists. The article asserts – quite boldly – that the exponential development of machine intelligence in legal services is inevitable and that the application of machine intelligence to legal services will result in better quality legal services at a lower cost. It posits that opposition to machine intelligence comes from fear of machine intelligence or fear of the implications for lawyers’ monopoly. The article describes the first “fear” as lack of trust regarding the competence of machine intelligence or concerns about the inability of the legal profession’s competence to evaluate machine intelligence. As to the second, the legal profession’s monopoly is implicated because if you accept the article’s starting premises, which are listed above, then the rise of machine intelligence might lead the public to believe that the current monopoly is unnecessary as a matter of “expertise” and that it impedes rather than promotes the public good because it hinders a broader more cost-effective distribution of the competent legal services.
In my view, the third section of The Great Disruption would have been stronger if it had addressed more directly the competing visions of machine intelligence. For example, it might have been helpful if this section had labelled some of the potential reactions as “skepticism” rather than “fear.” While I agree with the authors regarding the likely future abilities of machine intelligence, I know that others are more skeptical about whether machine intelligence can deliver what has been promised. Some may believe that lawyers are fundamentally more suited than machines to provide the customization that distinguishes legal services and legal advice from legal information. Labelling these reactions as “skepticism” rather than “fear” might have promoted greater dialogue about the nature of legal services and the capabilities of machine intelligence.
The third section of the article would also have been stronger if it had addressed the perspectives of those who think that machine intelligence might turn out to have results that are less beneficial than the rosy picture described by the authors. For example, this section might have referred to concerns about the lack of transparency and accountability that can occur when the values and assumptions that underlie machine intelligence algorithms are not clear or concerns that machine intelligence algorithms may be “gamed” by those with the most experience or power. These types of issues have been discussed in the context of online dispute resolution. In defense of the article, however, while an expanded discussion of these kinds of issues would have been useful, the article was subject to strict word limits and it clearly has laid the groundwork for ongoing dialogue.
In sum, not all readers will agree with the article’s concluding sentence, which states: “Ultimately, therefore, the disruptive effect of machine intelligence will trigger the end of lawyers’ monopoly and provide a benefit to society and clients as legal services become more transparent and affordable to consumers, and access to justice thereby becomes more widely available.” All readers should agree, however, that this is a thought-provoking article that deserves serious study. The implications of the developments the authors discuss are profound. While I am agnostic about whether machine intelligence ultimately will be deployed in a way that helps the public or whether it has the potential to be harmful, I am not agnostic with respect to the significance of these developments. Professors McGinnis and Pearce have convinced me that dramatic changes inevitably will occur in the very near future. In my view, it would be a colossal mistake if lawyers – and those who teach aspiring lawyers – ignore the issues implicated by Professor McGinnis and Pearce’s article.