Google DeepMind, the artificial intelligence lab owned by Google, has launched the DeepMind Institute, a forum where its researchers and outside experts can examine artificial general intelligence and its effects on society.
Artificial general intelligence, or AGI, refers to systems that match or exceed human ability across most cognitive tasks.
The institute is led by Shane Legg, a DeepMind co-founder, with senior involvement from the chief executive, Demis Hassabis, and James Manyika, a senior Google executive.
It opened with papers on economic policy for AGI, the transparency of model reasoning, global access to the technology and human flourishing.
"Humanity has to be a part of this," Manyika said, framing the institute's remit as global rather than tied to any one country.
The launch follows a period in which Google consolidated its frontier AI research and shifted some emphasis from building models internally towards public work on governance, safety and equitable access.
DeepMind has publicly signalled that AGI may be near, and the institute appears designed to broaden interdisciplinary debate about how to manage that transition.
