"Computers will, in the not so far away future, be able to replace the human Judge, at least for straight forward cases”. This is the conclusion of the session “Robot vs. Judge” on the “Dag van de Rechtspraak” (Jurisdiction Day) in the Utrecht Tivoli Vredenburg last week.
In this session, Prof. Dr. H.J. Van den Herik of the Leiden Center of Data Science (LCDS) and Prof Dr. Ir. J.C. Scholtes of Maastricht University / ZyLAB, talked about the influence of technology and Artificial Intelligence on the future role of the judge, and their answer to the question “Can computers adjudicate?” is a strong yes, certainly for simple cases, and in the foreseeable future probably also for more complex cases.
The basis for their conclusion is the assumption that human law is based on rules, values, case based reasoning and data science. Exactly this is the field where computers outsmart humans. The professors illustrate this with the world famous examples of the machines beating the best players of chest and Jeopardy and of course with the latest breakthrough in February of this year, when Google's DeepMind AI system demonstrated an ability to learn independently from its own memory, show intuitive behavior and beat the world's best Go players.
They foresee a technology-driven society where human and machine intelligence are interwoven. The challenge lies in finding a solid balance. Because only if we can formulate the law in an exact an unambiguous way, computers will be able to litigate in a neutral way. We are not there yet. As computers become more “intelligent,” some data scientists have been puzzled as they’ve observed their algorithms being sexist or racist. This should not be surprising, these algorithms were trained with social data that reflect society’s biases, and algorithms amplify these biases to improve their performance metrics.
So no machine judges yet, but today’s business reality is very simple: we need technology. In today’s world of information overload, manual search is obsolete. There simply is too much and too complex data. You never know exactly what and where to search and you can never be sure that you do not miss relevant data. We need legal technology in many ways.
At the end of the session, Prof Scholtes showed how he loaded all available Dutch case law from www.rechtspraak.nl in the ZyLAB ONE system and how he text-mined, used advanced search techniques, clustering and even machine learning to find relevant case law easier, faster and better. Finding such information is a first step which computers can already do better than humans. Extracting relevant information and reason with these is the next, we still need to work on this, but we are making significant steps in the right direction as we speak.
It will take a few more years, but the robot as judge is definitely in sight.
Please see previous blogs for more reading on how Artificial Intelligence technology can help in eDiscovery, Regulatory Requests or Internal Investigations or join us for the first of a series of two webinar on AI based analytics.