A neural network is bypassed professional lawyers in the competition for the interpretation of documents

Techno 27 February, 2018

2018-02-27 15:10

A neural network is bypassed professional lawyers in the competition for the interpretation of documents
Best lawyers in America wrestled with artificial intelligence and find out which of them faster and better cope with the interpretation of legal documents. In the end, the robotic system proved to be more savvy than experienced professionals.

Sign up for news “UkrMedia” in Facebook, Twitter or Google+

Competition with an experimental character has passed on the legal online platform LawGeex under the supervision of professors from Stanford University, school of law Duke University and the University of southern California. For the right “best” fighting 20 highly qualified lawyers and a specially trained system. In four hours the participants had to learn five agreements on nondisclosure of information and to identify 30 disorders, some of which required the arbitral Tribunal and damages. They are evaluated by how accurately they could identify each of the cases, reports Rus.Media.

As a result, the neural network has achieved 95 percent accuracy in the interpretation of the various clauses of the agreement, while the accuracy of experts was 85%. Moreover, artificial intelligence has coped with the task in just 26 seconds, and the lawyers it took almost two hours – 92 minutes. The highest computer – 100% accuracy in one issue, a person he did not exceed 97%.

Grant Gulovsen (Grant Gulovsen), intellectual property expert and one of the participants said that their tasks were very similar to those which he and his colleagues face on a daily basis. By the way, he admitted that he does not feel vanquished, as it believes that machine learning in the future will facilitate the work of lawyers and give them the opportunity to focus on those cases that will always require a “human” approach.

Agree with him and Erica Buell (Erika Buell), one of the initiators of the experiment, Professor at Duke University. She is confident that the car will be able to perform monotonous work, but advice as before will hold only people. “I think law students and young lawyers need to understand how to construct artificial intelligence to use his ability to its activities, says Buell. – And customers will be delighted, if with their computer problems will be solved faster.”