DeepMind CEO Demis Hassabis says AGI may arrive by 2029, warns world is unprepared
Google DeepMind CEO Demis Hassabis warned that artificial general intelligence (AGI) could arrive as early as 2029, faster than previously expected, and urged governments to accelerate AI safety regulations. He highlighted rapid advancements in AI agents and autonomous systems as signs of an impending 'agentic era,' stressing society remains unprepared for potential disruptions and risks like recursive self-improvement.
Google DeepMind CEO Demis Hassabis issued a stark warning about the potential arrival of artificial general intelligence (AGI) by 2029, earlier than previously anticipated. Speaking after Google’s annual developer conference, Hassabis noted recent breakthroughs in AI agents and autonomous systems suggest the industry may be on track for AGI sooner than expected, though he still considers 2030 a likely milestone. Hassabis described the current 'agentic era'—where AI systems perform tasks independently—as a preview of future disruptions. He pointed to AI agents improving productivity in software development and automation, signaling how advanced systems could soon surpass human capabilities in reasoning and planning. The pace of progress has already surprised businesses and governments, with incidents involving AI models from companies like Anthropic serving as a 'warning shot' for broader unpreparedness. Early signs of 'soft self-improvement' in AI, such as coding assistants accelerating development, raise concerns about recursive self-improvement, where AI could drive its own advancement. Hassabis acknowledged that leading labs are monitoring this closely due to both its potential benefits and risks. He called for faster government action on AI safety and regulation, arguing that policymakers and economists underestimate the speed of AI progress. While he supports discussions on stricter oversight, including mandatory safety testing before model releases, he believes current efforts are too slow. Hassabis noted that major AI companies are already collaborating on safety frameworks, but the next few years will determine whether societies adapt responsibly or face cycles of excitement and backlash. With AI capabilities advancing rapidly, Hassabis emphasized that the world has a narrow window to prepare. He warned that delays in addressing safety and regulation could leave societies vulnerable to unforeseen consequences as AGI approaches.
This content was automatically generated and/or translated by AI. It may contain inaccuracies. Please refer to the original sources for verification.