Elon Musk, the maverick entrepreneur behind Tesla and SpaceX has made yet another of his predictions. He says that Artificial Intelligence (AI) will be superior to humans in five years.
His prediction is also a warning. Musk has been outspoken about the dangers of AI in the past: in 2018 he claimed AI could become “an immortal dictator from which we would never escape” and even said he thought the technology was more dangerous than nuclear weapons.
Ryan Daws writing for AI News, refers to a recent New York Times interview given by Musk in which he said current trends in AI suggested it would “overtake” humans by 2025. Musk then added, ““that doesn’t mean that everything goes to hell in five years. It just means that things get unstable or weird.” Well, we’ve already experienced that in 2020 and it had nothing to do with AI.
Ray Kurzweil, an eminent futurist has previously estimated that machine intelligence overtaking human intelligence would occur around 2045.
It is ironic perhaps that Musk’s companies are all heavy users of AI, but as Daws says, it isn’t the case that Musk is against the technology: he simply thinks it should be more regulated – an ethical AI if you like.
Indeed, Musk formed OpenAI in 2015 to research and promote ethical artificial intelligence, although he left it in 2018 due to internal disagreements. In February this year, he said that OpenAI should be more “open” and that all organisations “developing advanced AI should be regulated, including Tesla.”
AI and fake news
It was interesting to read in Daws’ article that OpenAI had developed a text generator but decided not to release it, citing its dangers in a word that is already really struggling with the surge in fake news. However, two graduates created something similar, claiming it “allows everyone to have an important conversation about security, and researchers to help secure against future potential abuses.”
Now OPenAI has allowed a select band of researchers to try out its AI text tech. Called GPT-3, it has been grabbing attention due to the “incredible things it can achieve with limited input.” For example, one researcher tweeted: “Playing with GPT-3 feels like seeing the future. I’ve gotten it to write songs, stories, press releases, guitar tabs, interviews, essays, technical manuals. It’s shockingly good.”
That all sounds very exciting, although song writers and PR agencies may not feel the same level of thrill about it, until they discover how much easier it makes their work. Will human intelligence be overtaken by 2025? It’s more likely that Musk’s prediction is simply attention grabbing, something that he excels at. Perhaps he’s wondering if AI might overtake his ability to stay in the headlines?