Couple years ago, there were no computers, no internet, no robots. Nowadays there is a competitor of the human intelligence, even though developed by people. It is called artificial intelligence and it helps people in plenty of different areas, including law. “I want computers to assist judges and attorneys to make their own decisions as effectively and efficiently as possible,” said Professor Kevin Ashley from University of Pittsburgh, expert on application of artificial intelligence in law.
How can artificial intelligence be used in practical life of people?
Artificial intelligence is a subarea of computer science where researchers try to get computers to behave in a manner we call intelligent when performed by humans. In effect, the researchers build computational models of intelligent behavior. One way that artificial intelligence can help people in practical life is to improve the interfaces to other kinds of computer tools. For instance, an intelligent “front end” to an information retrieval system can help ensure that, of all the materials retrieved, those that are most relevant to solving the user’s particular problem are highlighted.
And how was artificial intelligence used in practical life so far?
Artificial intelligence has had many diverse applications, for example, to make it easier for citizens to fill out their income tax returns, to teach algebra and geometry to grade school students, to help applicants and civil servants navigate a country’s immigration laws and procedures, to automatically summarize the main points of articles in newspapers, and to automatically translate texts from one language to another.
How can artificial intelligence be applied in the domain of law?
It is connected by researchers who build computational models of legal reasoning. That is, they study the intelligent behavior of legal reasoners, I mean judges, attorneys and law students, and build computer programs that can approximate such behavior. Today, researchers in the field of artificial intelligence and Law have built computer programs that can reason with legal rules and precedents, make arguments like a legal advocate, and predict outcomes of new cases. These programs tend to have limited areas of application; they may only be able to deal with tens of rules and a few hundred cases in some particular area of law, but their outputs are comparable to those of human reasoners. The challenge is how to connect the computational models more generally to the tasks that attorneys, judges, and law students actually perform in daily professional life and to numerous complex texts they actually use.
How is it possible that computer can understand textual rules or answer textual questions?
Computers have only a very limited ability to “read” and understand complex texts such as statutes. Using natural language processing and machine learning, however, useful information can be extracted from statutory texts, e.g., the type of law like Administrative Law, Private Law, Computer Science Law etc., functions of a regulatory provision (e.g., definition, liability, prohibition, duty, permission, or penalty), features related to those functions (e.g., bearer of the duty, action, counterpart, object), and a kind of profile useful for retrieving relevant provisions from statutory databases.
Can we expect computers to replace human judges in the future?
It is not my goal to see computers ever replace judges and attorneys. Instead, I want computers ever more intelligently to assist judges and attorneys to make their own decisions as effectively and efficiently as possible. This requires the computer program to have a model of the judge’s or attorney’s decision-making process and be able to find legal and other information in databases that is relevant to that process and to present it in a way that informs the decision-maker about its relevance and utility. Legal information systems do something like this today, but by applying artificial intelligence, we can greatly extend their capabilities to assist judges, attorneys and also law students!