During the workshop Jaap van den Herik was talking about AI (Artificial Intelligence) and the question whether computers are able to replace people. He used court cases as an example to show what computers can do. Some problems for judges and computers in a court case are the use of vague concepts, norms, values and open textures.
His research showed that computers are able to search and find revealing information and new precedents. They are also able to create things by automatically filling in forms and writing verdicts as an autonomous unit. Another thing computers are capable of is using legal analytics to predict the outcome of a casus. He concludes that most of the work in a court case can be done by a computer.
Van den Herik generalizes this example with the statement: “Computers will surpass human beings in all sorts of intelligence tasks”. He wonders what will happen when this comes true. Other scientists, like Stephen Hawking, think that this shall threaten the existence of mankind.
The solution of Van den Herik is to use safeguards in systems based on artificial intelligence. Possible safeguards would be to diminish or augment the attention to AI. Both options are undesirable, but for different reasons. The best solution might be to introduce safeguards like moral constraints in AI systems, creating ethical committees or paying more attention to support systems which would reduce the consequences of hacking an AI-based system.
By Amanda Schaaf