Is Elon Musk Scared of AI? SpaceX Founder Urges Tech Company to Develop 'Symbiosis' Between Man and Machine

elonmusk_ai_dota2
Elon Musk speaks during a June news conference in Chicago. Musk, a known artificial intelligence skeptic, urged a tech company to develop brain-controlled computers that foster cooperation between man and machine rather than one's domination over the other. (Photo by Joshua Lot/Getty Images)

After a crew of artificial intelligence bots bested top-of-the-league humans in a video game match on Sunday, Elon Musk urged OpenAI, the research company he co-founded that created the machine team, to next explore ways to forge "symbiosis" between man and machine rather than complete domination.

The best-of-three match of the online battle arena game Dota 2 pitted five humans who ranked in the top 99.95 percentile of players against a team of five neural networks known as the OpenAI Five, which learned the game by playing the equivalent of 180 years per day against itself. In two out of the three matches, the bots clobbered the humans, who won the final game only after the audience chose the machines' players and purposely spoiled their chances of a sweep.

After the match, staunch AI skeptic Musk tweeted the research team had done "great work" but should introduce brain-controlled computer devices that allow humans to work with AI instead of against it.

Musk, the founder of SpaceX and co-founder of Tesla, has been slow to accept the onslaught of machine intelligence, a contrast to other entrepreneurs who have widely adopted the technology. In March, he told South by Southwest audiences that AI's development will soon outpace that of humans and could lead to an "advent of digital superintelligence" if research goes unregulated.

"I think the danger of AI is much greater than the danger of nuclear warheads by a lot, and nobody wants to suggest that we allow anyone to build nuclear warheads if they want," he said. "That would be insane."

His statements warning of a digital apocalypse are numerous: In an interview for the documentary Do You Trust This Computer? Musk said a "godlike" AI agent could "take over the world" in a robot dictatorship. He's also claimed competition between global superpowers for AI superiority is the most plausible cause for World War III and that tech is a more formidable threat than North Korea.

Despite his marked incredulity of the technology, he started his own AI firm in 2015 to research ways it could "benefit humanity as a whole." He resigned from OpenAI's board in February to avoid potential conflict with his work at Tesla, where he and his team have been testing autonomous driving technology.

OpenAI heads said they chose to train their bots in the game because of its complexity and popularity to slightly even the playing field between the machines and human players. Researchers imposed limitations on the game for their incipient experiments, narrowing character selection to just 18 out of more than 100 possible heroes for both human and bot players and allowing both teams to use strategies and gameplay moves that bettered their chances.

The OpenAI Five, so far undefeated in best-of matches, will next play professional Dota 2 teams at the International, the biggest tournament for the game where the world's best teams compete, in late August. The prize pool is currently valued at $23.5 million. In an exhibition match at the International in 2017, a single OpenAI bot beat one of the best Dota 2 players in the world.