Google Develops Advance AI Training Model through DeepMind

The company reveals that Google will use DeepMind's new JEST, Joint Example Selection, in 2024. The new method includes less usage of conventional artificial intelligence. It depends on conventional training methods that focus on individual data points for training and experience. However, JEST promises to reduce the technological sector's training time and energy consumption. Research experiences on Google's DeepMind and its implementation of JEST show that it is ten times more efficient and 13 times faster than previous ones. Hence, this development in AI training by Google will have different applications in the industry. Further, it will increase the pace of production and training of AI models in 2024.

According to the experts, the JEST method by Google DeepMind follows its new way of training. It is different from the conventional AI training methods. JEST focuses on the data subsets that are most learnable rather than focusing on the individual data points. Hence, the new method includes building smaller models for data evaluation that can be a guide for larger models. JEST anticipates improving AI applications in finance, healthcare, education, and others. Experts hope Google's DeepMind will set new limits with its finest AI models and training. Hence, the tech sector believes this latest development will revolutionize the artificial intelligence field in the industry.

                               Google’s AI training examples recorded by the study: