SoftBank’s Son Says Artificial General Intelligence Will Soon Surpass Humans

livemint.com:

TOKYO—SoftBank Group’s chief executive officer, Masayoshi Son, said artificial general intelligence is so powerful that within a decade it will surpass all human knowledge. “Take advantage of it or be left behind," Son said to a crowd gathered for the technology investment company’s annual SoftBank World event.

Companies from Meta Patforms to Google parent Alphabet to ChatGPT’s developer, OpenAI, have been racing to be at the forefront of artificial intelligence, which proponents have said will transform the way people work, play and live. Son, 66, has made it clear he is a believer.

He spent much of his roughly hourlong address Wednesday arguing for widespread development and adoption of artificial general intelligence, a computer system able to match human intelligence and reasoning. He said he had been talking to GPT-4, a more advanced version of the technology behind the artificial intelligence chatbot ChatGPT, every day to debate ideas about potential inventions.

When he grew tired of debating GPT-4, he created characters on the chatbot and made them debate each other, serving as “judge" and noting the merits and weaknesses of different arguments. By 2030, Son said he believes artificial general intelligence will be 10 times more powerful than all of human intelligence.

He laid out a road map of what he believes will be achieved in the decades that follow: a full transition to self-driving vehicles that will nearly eliminate road accidents, and Nobel Prize-worthy advancements in engineering and science generated by artificial general intelligence alone. With the screen behind him showing a photo of a goldfish, trapped in a bowl and looking at a question mark, he made it clear what he thought the audience would