Experts from T-Bank AI Research and the AIRI Institute have created the world’s first contextual learning model, enabling artificial intelligence (AI) systems to independently learn new actions and make decisions using just a few examples. Previously, for AI to perform new tasks, it had to be completely retrained, which was an expensive and time-consuming process. Now, thanks to Russian specialists, AI will be able to adapt to changes in the external environment and new tasks without human intervention. How the Contextual Learning Model Works This model is a database that includes over 100 billion sets of information. Unlike random data, such as text-based questions and answers, the database contains records of AI actions in specific situations, allowing the system to learn and adapt based on real experience. Application of the Scientists’ Discovery For example, it will be useful in the development of assistant robots, which will be able to adapt to the unique conditions of each home. In the field of autonomous vehicles, this technology will allow cars to continue operating even after a part is replaced that functions differently, without the need for complete reprogramming. In today’s world, artificial intelligence is used in almost all areas of life. With its help, various works of art can be created, buildings designed, diseases diagnosed, and even terrorist attacks prevented. For example, the Russian company Smart Engines previously introduced its own development capable of finding criminals, including terrorists, among migrant workers.