Artificial intelligence (AI) is clearly ubiquitous today. It is used to support or replace humans in difficult analytical and decision-making applications in areas such as banking, e-commerce and tourism. Examples of AI-assisted systems in these areas are respectively the robo advisors that provide investment advice, engines that recommend products to buy, and chatbots that help book flights.
AI systems have also recently entered the arena of professional sports, as evidenced by the Japanese company Fujitsu which is developing an AI-assisted judging system to assess the performance of athletes in artistic gymnastics. Judging in artistic gymnastics is a complex and laborious process, which requires significant physical and cognitive effort from the judges, a high level of concentration during the competition and a high level of education on the code of points.
Impact on evaluation
In two weeks, the World Gymnastics Championship starts in Liverpool, a good opportunity to pay close attention to the judging process. According to the opinions of athletes and judges, evaluation in competitive sports is subject to various types of human biases, errors and inaccuracies, which impact the overall evaluation of athlete performance, which can have dramatic consequences on the medal chances and career progression of gymnasts. In addition, international gymnastics competitions may last for long hours a day, and judges may feel tired, hungry, have an unfavorable viewing angle, have light in their eyes, etc., which may also affect the overall quality of the judging process.
For this reason, several artistic gymnastics stakeholders anticipate the arrival of an AI-assisted system that should solve, or at least partially compensate for, the current problems of the human system, such as the lack of precision , consistency, objectivity and transparency in the judging process. Combining 3D laser cameras to capture gymnasts’ movements and AI that matches movements to the point code, the system is intended to make more accurate, fairer, more transparent and less biased judgments in real time. .
challenges and problems
In this context, we formed an international group of researchers from the University of Liège (Belgium), the University of Aalto (Finland) and the UNSW Business School of Australia in order to discover the positive and negative implications the introduction of an AI-assisted system for humans, organizations and society
. We have found that the introduction of AI can lead to new challenges and problems, some of which may even be paradoxical. For example, AI can effectively make the evaluation process more transparent by explaining how a score is obtained. At the same time, some of the correspondence that occurs inside the system remains a “black box”. Similarly, the AI-assisted system may eliminate some human biases (eg, order effect), but it also introduces new biases (eg, based on body type). Also, the very “soul” of artistic gymnastics revolves around artistry, but how can AI judge that? Finally, what if Nina Derwael or Simona Biles comes up with another never-before-seen routine, will the AI just reject it?
In summary, the complex, uncertain and paradoxical nature of AI-assisted systems shows that it is strongly recommended to exercise some caution in the process of their implementation. And in the pursuit of increased productivity, efficiency, cost-effectiveness, precision, reliability and speed of processes, it is also necessary to remember the various possible risks and disadvantages. that these systems can bring to the organization and to society.
Furthermore, our research results show that it is important to consider a hybrid cooperation between human professionals and AI; when, instead of being adversaries, AI and humans should team up to function as an integrated unit and augment each other’s capabilities. If done correctly, it would achieve the best of both worlds.Mazurova E., Standaert W., Penttinen E., & Tan F. (2022) “
Paradoxical Tensions Related to AI-Powered Evaluation Systems in Competitive Sports,” Information Systems Frontiers, Vol 24: 3 pp. 897 – 922. (
We would love to give thanks to the writer of this short article for this remarkable content
Artificial intelligence to judge sports competitions?
Visit our social media profiles along with other pages related to themhttps://www.ai-magazine.com/related-pages/