Sir Demis Hassabis, the head of Google DeepMind, has stressed the urgent need for more research into the threats posed by artificial intelligence. Speaking exclusively at the AI Impact Summit in Delhi, Hassabis stated that the industry requires "smart regulation" to address the "real risks" associated with AI.
Many leaders at the summit have pushed for increased global governance of AI. However, the United States has opposed this, with White House technology advisor Michael Kratsios asserting that AI adoption should not be hindered by "bureaucracies and centralised control."
Hassabis highlighted two primary concerns: the potential misuse of AI by "bad actors" and the risk of losing control over increasingly powerful autonomous systems. He acknowledged that regulators face challenges keeping pace with AI development.
While Sir Demis believes the West holds a slight lead over China in AI development, he cautioned that this gap could narrow within months. He emphasized the responsibility of balancing bold AI deployment with robust safety measures, stating, "We don't always get things right, but we get it more correct than most."
Looking ahead, Hassabis predicts AI will become a "superpower" in creation within the next decade, underscoring the continued importance of STEM education for navigating these advanced systems.