Google to deploy AI to detect personal crises via search queries

Google will update its machine-learning search model MUM to detect a wide range of crisis-related search queries concerning domestic violence, substance abuse, and suicide. The Verge reports.

According to company representatives, users frequently turn to Google Search to resolve personal problems. In such cases, MUM will be able to detect queries that earlier search tools could not recognise, said Anne Merritt, the company’s product manager for health and information quality.

\”MUM can help us understand longer or more complex questions, such as ‘why he attacked me when I said that I don’t love him’\”, she said.

Without AI, systems find it difficult to understand such long natural-language queries, Merritt added.

Google expects that the update to the search engine will provide users with more useful and timely information when they need it.

In the coming months, with the help of local partners, the company plans to use AI to improve crisis-search capabilities in other countries, as MUM supports operation in 75 languages.

Google also said it plans to deploy the model in other products, including to improve spam protection and strengthen security measures in countries where it has little training data.

Earlier, at the I/O developer conference in May 2021, the tech giant introduced the MUM language model.

In January 2022, Google added automatic translation of subtitles to the Meet video-conferencing service.

Subscribe to ForkLog’s news on Telegram: ForkLog AI — all the news from the world of AI!