DeepMindCo-FounderDemis Hassabis recently said in a podcast that the currentAIHowever, he also stressed that the long-term potential of the technology, especially in terms of general artificial intelligence (AGI), is still underestimated.
Hassabis described the public's attention to artificial intelligence as a double-edged sword. While widespread interest suggests that artificial intelligence will play a positive role in daily life, this attention also brings a lot of noise. For experts like Hassabis who have been deeply involved in the field of artificial intelligence for many years, it is both gratifying and surreal to see the public gradually realize the importance of artificial intelligence in the future.
Source Note: The image is generated by AI, and the image is authorized by Midjourney
Although Hassabis believes that current AI models are overhyped, he praises their ability to infer real-world information through language learning alone. He noted: "It's very surprising that these models learn in an abstract domain and yet are able to infer information relevant to the real world."
Hassabis further said that people have too high expectations for the short-term capabilities of artificial intelligence. He mentioned that many people and startups claim that this technology can achieve goals that the current state of the art has not yet achieved, and called these claims "crazy ideas" that venture capitalists are still chasing even though the underlying technology does not exist yet.
However, Hassabis believes that the long-term potential of artificial intelligence is still not fully realized, especially in the field of AGI. He predicts that AGI may be achieved within the next decade and believes that this technology will bring huge potential benefits to mankind, including curing diseases, developing clean energy and combating climate change. DeepMind is designed as a 20-year project and is expected to be completed around 2030.
Hassabis also stressed that as AI systems continue to grow, it is critical to prevent the risk of human abuse and loss of control. He called on the international community to strengthen cooperation to regulate AI, and predicted that the technology would be difficult to effectively control within a country. In the final stages of AGI development, he envisioned that the world's top AI researchers would form a scientific "Avengers" team to jointly meet this challenge.