Google and DeepMind Launch AI Safety Institute Amid Global Cooperation Calls

Alphabet Inc.'s Google and DeepMind have launched the DeepMind Institute to explore the potential impact of artificial general intelligence (AGI) on society. The institute's first publications cover topics such as economic policy, model reasoning transparency, global access, and human flourishing.

The launch comes amid growing calls for international cooperation on AI, with its leaders emphasizing that addressing safety and responsibility cannot be country-specific due to the global nature of AI advancements.

Shane Legg, managing editor of the DeepMind Institute, alongside Google senior vice president James Manyika and DeepMind chair Demis Hassabis, aim to make complex research accessible and encourage broader debate about AGI's future. They acknowledge that researchers may disagree as new data emerges and views evolve.