EXPERT SYSTEMS: FROM LEIBNIZ’ DREAM TO THE ROBOT LAWYER In 1950, mathematician Alan Turing published a famous paper under the title "Computing Machinery and Intelligence". It started with a key question: Can machines think? That text was the foundation of the field of artificial intelligence. Artificial intelligence is a branch of computer science that develops systems to perform typically human tasks such as voice recognition, understanding language, problem solving and decision-making. In the 1950s and 1960s, during the early days, pioneers like Herbert Simon and Marvin Minsky made grandiose predictions. In a few years, they said, machines would be able to perform almost any human activity. Those ideas were reflected in the comics, the books and the science fiction movies of that time. But the actual progress was way more modest. So in the 1970s, the AI field fell into oblivion. Those years were known as “the winter of artificial intelligence”. In the 1980s, there was renewed interest in AI with the creation of expert systems. These are decision trees that mimic the reasoning of human experts. For example, the reasoning followed by a doctor to diagnose a disease. Once the process has been systematized, it can be done by a non-expert or by a machine. In the field of law, people tried to pack legal reasoning under the form of software. For example, automating the drafting of a will through a series of questions. Are you married? Do you have children? Which part do you want to leave to each child? After the information has been collected, the program drafts the document. Rocket Lawyer and LegalZoom, two companies we mentioned in the previous module, are based on expert systems for document drafting. The user follows a series of steps and the software creates a document. By following rules, these programs allow non-expert users to draft contracts. A major milestone in artificial intelligence occurred in 1996. That year, IBM’s Deep Blue software defeated the world chess champion Gary Kasparov. In 2011, another IBM program called Watson defeated the best players at the Q&A game Jeopardy! That game showed to the world the progress of AI in the previous 15 years. Deep Blue could only play chess. Watson could understand questions in natural language, search the answer in a database and reply with a human voice. Born in 2014, the company ROSS Intelligence applies Watson technology in the legal industry. The user asks a question in natural language and ROSS seeks the answer in a large database of cases and legislation. Thus, it allows to quickly find information to prepare a case. At the moment, law firms can use ROSS to automate their legal research processes. But we can already foresee where this technology could be in a few years. A company could pay a monthly fee for a virtual lawyer service. For a small fee, you could ask all kinds of questions and get an immediate answer. Under Argentinian law, how long a notice one has to give an employee before firing him? How should severance be calculated? Instead of asking a lawyer, many issues could be solved with software. Chatbots are another field of application of expert legal systems. A chatbot is a program that interacts with the user following the rules of a decision tree. You probably have already interacted with these robots in customer service. Now they are breaking into the legal industry. In 2018, Stanford University student Joshua Browder developed the DoNotPay application to automate the appeal of traffic fines. If you got a ticket, you download the app and a robot lawyer guides you through a series of questions. At the end, it drafts a document to be presented at the traffic authority. DoNotPay started with fines and then extended its applications to small consumer claims, e-commerce, airlines and banks. Expert systems were the first step of the AI disruption in the legal industry. But in recent years, a new AI technology made all of that progress seem like child's play. Machine learning brought a number of applications that might fundamentally transform the practice of law. We will discuss this in the next lesson.