Geoffrey Hinton, the renowned ‘Godfather of AI’ who left Google last year to warn about the technology’s dangers, has made a startling prediction: OpenAI’s GPT-5 could achieve artificial general intelligence (AGI) as early as 2025. In a recent interview with The Times, Hinton suggested that the next iteration of OpenAI’s large language model might surpass human intelligence in critical ways, marking a significant milestone in AI development that’s arriving faster than many anticipated.
Hinton’s prediction represents a dramatic acceleration of his previous timeline, having earlier suggested AGI might emerge by 2027. This shift reflects the rapid pace of advancement in AI capabilities, particularly in reasoning and problem-solving domains traditionally considered uniquely human. The AI pioneer specifically highlighted GPT-5’s potential to exceed human intelligence in areas like coding and mathematical reasoning, while acknowledging that full human-level intelligence across all domains remains a more distant goal.
As one of the most respected voices in artificial intelligence, Hinton’s warnings carry significant weight in the tech community. His concerns about AI’s potential dangers prompted his departure from Google in 2023, and his latest timeline prediction adds urgency to ongoing discussions about AI safety and regulation. With OpenAI reportedly working on GPT-5 despite earlier denials, the AI industry appears to be racing toward a future where machines may soon rival or exceed human cognitive abilities in specific domains, raising profound questions about the future relationship between humans and increasingly intelligent machines.
Source: https://www.businessinsider.com/geoffrey-hinton-godfather-ai-gpt4-openai-2025