Attention, tech industry: Elon Musk’s artificial intelligence can beat up your artificial intelligence. And based on the whooping it recently gave professional eSports gamers, it can beat your human intelligence as well.
But this is more than just another instance of humans creating a machine that’s really good at one specific, arbitrary game (like AlphaGo defeating the Go world champion or IBM supercomputer Deep Blue defeating chess grandmaster Garry Kasparov).
According to the SpaceX founder and Tesla CEO, his new A.I. is more than just a party trick, and could instead be the secret weapon to winning the next World War.
How Open A.I. Beat Dota2
The game Musk’s artificial intelligence research company (OpenA.I.) showed their work off on was Dota2.
Infinitely more complex than even chess, Dota is arguably the most competitive and popular eSports game right now, regularly drawing live crowds of 15,000+ and over five million online viewers during its annual championship event, The International.
During the 2017 iteration of the tournament, a 1v1 demonstration match was held between OpenA.I.’s machine-learning bot and several professional Dota players. Not only did the bot dominate and go undefeated against several different opponents, it even made two-time International runner-up Dendi quit in the middle of the second game out of frustration.
How the bot got so good at the game isn’t that different compared to how humans do, through practice.
A tiny difference is that the bot took just two weeks to get to a professional level of playing, compared to several years or more.
Playing against a copy of itself, the bot self-learned small improvements it could make in each match, quickly going from random movements to elite-level skills.
The OpenA.I. team hopes that this is just the beginning and that they can soon create bots that can take down a team of humans or learn to play cooperatively in a mixed human-machine game (typical Dota matches are 5v5).
An A.I. Fought World War
As with most things Elon does, he sees this as a small victory toward a much larger goal: developing A.I. that has a positive long-term human impact.
But perhaps the biggest irony of his progress toward this goal is that Musk also views it as the single biggest threat to humanity.
In a September 2017 rant, Musk went on for several Tweets about how “competition for A.I. superiority at a national level will most likely be the cause of WW3”.
The somber thought was prompted by a recent speech from Russian president Vladimir Putin in which Putin said, “Artificial intelligence is the future, not only for Russia, but for all humankind,” and “Whoever becomes the leader in this sphere will become the ruler of the world.”
Putin has also predicted that future wars could be fought entirely by drones, with the spoils going to the nation(s) with the most still standing (or flying).
Months earlier, Musk asked an assembly of American governors to be more proactive than reactionary in regulating A.I.
Musk said, “Normally the way regulations are set up is a whole bunch of bad things happen, there’s a public outcry and then after many years the regulatory agencies set up to regulate that industry. A.I. is a fundamental risk to the existence of civilization in a way that car accidents, airplane crashes, faulty drugs or bad food were not.”
Musk went carried on, taking to Twitter saying that a World War III triggered by A.I. might not be a case of a country deliberately telling it to attack another. Instead, he fears it could be automated action by the intelligence that decides “a preemptive strike is most probably (sic) path to victory.”
A Digital Space Race
Over the summer, China declared it has a plan in place to make the country the world’s foremost A.I. developers by 2030.
The plan is part of a billion-dollar investment to further China’s technological advancements in areas similar to what Musk’s many projects focus on. The city of Tianjin has even set aside 20 square kilometers specifically designated as an intelligence technology hub for companies.
While the US is considered the other favorite in this digital space race, proposed cuts to American A.I. research organizations have researchers worried.
Paul Scharre, the director of a Washington D.C. tech think tank says, “A lot of people take it for granted that the US builds the best tech in the world, and I think that’s a dangerous assumption to make.”
Government funding isn’t the be-all end-all for fueling this kind of research though, as indicated by the amount of money the Facebooks and Googles of the world are throwing at A.I. research.
Where We Go from Here
Beyond just how wars are fought, A.I. is predicted to affect nearly every aspect of our daily lives, including eSports and online gambling.
And as scary and devastating as a scenario like the one in the movie WarGames might be if it actually played out, A.I. developments haven’t been all fire and brimstone.
At the inaugural A.I. for Good Global Summit last June, agencies presented ways in which the technology could help achieve the United Nation’s Sustainable Development Goals, including creating cleaner water, more sustainable cities, and even ending hunger.
Despite how easy A.I. makes things look, whatever happens will be a frustrating battle going forward. Just ask gamers.