DeepMind scientist challenges claims of AGI arrival, cautions on AI safety

The AI industry is currently engaged in a heated debate over whether General Artificial Intelligence (AGI) has already been achieved. In response to NVIDIA CEO Jensen Huang's public praise for OpenAI's Astra and his assertion that AGI has arrived, Shane Legg, co-founder and chief AGI scientist at Google DeepMind, publicly refuted this claim in an interview, stating that it is too early to declare the arrival of AGI. Legg emphasized that OpenAI's definition of AGI, as outlined in its bylaws, requires a system that surpasses humans in most economically valuable tasks. While Astra demonstrates impressive capabilities, Legg remains skeptical about its ability to meet this benchmark. He estimates a 50% chance of achieving minimum AGI by 2028. DeepMind's definition of AGI is broader, encompassing all cognitive abilities of the human brain. Legg also warned of the rapid pace of AI advancement and the need for robust safety mechanisms. DeepMind has established the DeepMind Institute, led by Legg, Demis Hassabis, and James Manyika, to study AGI's impact on society.