The development of AI will bring problems that need to be thought about today
More and more experts are warning that, together with positive changes, artificial intelligence will bring us serious problems. The dark side of AI has long been under the scrutiny of scientists. In February of this year, a 100-page report was dedicated to her, created by the joint efforts of several large organizations in the field of cyber defense and programming - the EFF Electronic Frontiers Foundation, the Institute for the Future of Humanity, the Oxford Center for the Study of Existential Risk, the University of Cambridge, the Center for New American Security and OpenAI . In the future, AI is expected to be used by attackers to carry out larger and more sophisticated attacks:
According to the researchers, we should prepare for new attacks that will implement the improved ability to analyze human behavior, mood and beliefs based on the available data. “First and foremost, we need to recognize that AI is a great tool for manipulating the masses,” says Peter Eckersley, EFF's chief technical officer. “Therefore, we are faced with the task of developing counteraction algorithms.” Fake newsThe manipulation of human behavior is something that can undermine the ability of democracies to conduct fair public debate. According to Jack Clark, director of strategy and communications at the nonprofit research organization OpenAI, we will be faced with massive amounts of compelling fakes in the form of images and videos. This will lead to an increase in propaganda ideas in the community and a growing amount of false news. There is a critical link between computer security and the exploitation of AI for bad purposes: if the computers on which machine learning systems are deployed do not have the necessary protection, sooner or later disaster will happen. “This is why the IT industry needs new investments,” says Eckersley. “AI is a double-edged sword. Either it will take cybersecurity to the next level, or it will destroy everything that we have achieved. Our task is to use it for defense purposes to ensure the stable and safe operation of electronic devices. ” Innovation and DevelopmentAI researchers and engineers must be aware of cyberthreat trends and be aware of the dual-use challenge of their work. Experts are calling for a rethinking of the norms of research openness, including risk assessments, access levels and sharing mechanisms. These recommendations cause concern for Daniel Castro, director of the Data Innovation Center: "Such measures can slow down the development of AI, as they do not fit the innovation model that technology is successfully developing." In addition to his words, Mr. Castro claims that AI can be used for various purposes, including fraudulent ones. However, the number of people eager to turn machines against humans is greatly exaggerated. Regulation of issuesIT experts hope to more specifically define the policy for the development and management of AI structures, but agree that some technical issues should be postponed. “Some of the negotiations are best postponed until the system is widely adopted,” explains Peter Eckersley. - However, there is another side. If you are working with an ever-changing system and you know for sure that it will take years to implement the necessary precautions, it is best to start as early as possible. ” One of the characteristics of public policy is that it rarely responds to problems at an early stage. According to Ross Rustici, senior director of research at Cybereason: “If we could get the political community to delve into security issues, convince researchers to focus on the problems of technology adoption, and not on the fact of innovation, the world would change dramatically for the better. Unfortunately, as history shows, our hopes are unlikely to come true. Usually we analyze the consequences of scientific breakthroughs after something bad happens. ” The Topic of Article: The development of AI will bring problems that need to be thought about today. |
