AI that can match humans at any task will be here in five to 10 years, Google DeepMind CEO says

Core Viewpoint - The emergence of artificial general intelligence (AGI) is anticipated within the next five to ten years, according to Demis Hassabis, CEO of Google DeepMind, who emphasizes that current AI systems are still limited in their capabilities [2][3]. Group 1: Predictions on AGI Timeline - Demis Hassabis predicts that AGI will start to emerge in the next five to ten years, highlighting the need for further research to enhance AI capabilities [2][3]. - Other industry leaders have varying predictions, with Baidu's CEO Robin Li suggesting AGI is "more than 10 years away," contrasting with more optimistic views from others like Dario Amodei of Anthropic, who sees advancements in "the next two or three years" [4][5]. - Cisco's Chief Product Officer Jeetu Patel believes AGI could be evident as soon as 2025, indicating a more immediate timeline compared to Hassabis [6][7]. Group 2: Challenges in Achieving AGI - Hassabis identifies the primary challenge in achieving AGI as the need for AI systems to understand real-world context, which is more complex than performing tasks in controlled environments like games [8][9]. - The development of "multi-agent" AI systems is gaining traction, which could facilitate better communication and cooperation among AI agents, a necessary step towards achieving AGI [10][12]. - DeepMind's work on AI agents in games like "Starcraft" illustrates the progress being made in developing agent-based systems that can compete and cooperate, which is essential for real-world applications [11].