Google Unveils MUM AI to Tackle Complex Search Tasks
Google introduces MUM, a powerful new AI model designed to understand complex search queries that currently require multiple searches. The multimodal technology aims to provide expert-level answers by processing information across languages and formats.
Google unveils a new artificial intelligence technology called Multitask Unified Model, or MUM, to solve complex search problems that currently require multiple queries. While traditional search engines force users to issue an average of eight separate searches for complicated tasks, MUM acts more like a human expert by understanding the nuanced context of a single question.
The new model is one thousand times more powerful than the previous BERT framework and demonstrates advanced capabilities by both understanding and generating language. It trains simultaneously across 75 different languages and various tasks, which allows it to build a much more comprehensive understanding of information and world knowledge.
Unlike previous AI models, MUM is multimodal and processes information across both text and images, with plans to expand into video and audio in the future. This versatile approach ultimately means that users need far fewer searches to accomplish complex goals, bringing Google significantly closer to delivering expert-level guidance.