AI safety research should be conducted independently of commercial frontier labs

From the articleFormer Google DeepMind researcher says he resigned because 'AI has the potential to kill us all'Also in Two Google Deepmind researchers lend voices to the AI apocalypse.