Video games have long been the best place to test your skill, intelligence, and courage against others. They’re also the place where we learned just how brutal computer opponents can be.
Of course over time, we learned to “game” the games, so to speak. Computers that used pattern-based attacks were easy enough to manipulate, and often it was only reaction times and controls that kept us from dominating.
That was until more advanced programming entered the game, and it’s getting even more advanced by the minute.
Dubbed OpenAI Five, it not only beat the humans, but it also worked alongside them and adapted to their play styles during the exhibition. The humans in question were five players from OG, which won Valve’s 2018 International in the game.
After proving the capability of OpenAI Five’s deep learning, the company issued an open challenge. From April 18 through April 22, gamers are invited to play against this venerable computer opponent.
Meanwhile in January 2019, Google’s DeepMind AI lab built AlphaStar, an AI that beat two of the world’s best Starcraft II players.
You may remember DeepMind for its mastery over humans in chess (something IBM’s Deep Blue and pretty much any chess game with a computer long beat us in), shogi, and AlphaGo.
Having destroyed us in traditional board games hailed by experts for increasing every level of human intelligence, it’s no wonder it wants to face us on more modern battlefields.
Will the robot revolution many fear be televised on Twitch?
Pandering to Our Robot Overlords
It’s not just gaming – AI is learning faster than we do in nearly every conceivable function.
AI reads faster, writes with no breaks, sees better, and can even create art and music. In fact, the only reason we can do the vast majority of what we do these days is because of computer technology.
We built computer and the internet in a way that they’re basically just layers upon layers. Engineers coming out of college are working on much different layers than those working with over a decade of experience.
Even the people building the systems often spend their time digging through archives of code praying for notes. They do this to find out why something was done, because even if they understand the language, understanding another person’s implementation is a whole other beast.
It’s no wonder we don’t understand all the data available to us, and we’re only adding more. While we lack understanding, deep learning and machine learning teams are training AI with buttloads of data.
So how do we coexist with something so much smarter than we are?
The answer is we already are, and we have been for a long time. We’ve always been slaves to data, and it’s ultimately the data that computers, AI or not, are reading and reacting to.
Computers only do what we train them to do, and that’s no different with AI. Even though machines do learn, the OpenAI Five can’t beat those five people in anything except Dota 2 right now. And DeepMind deploys different AI protocols to beat humans in each game.
It’s not any one individual AI that we have to fear – it’s whoever is at the helm of all of this technology.
And, like most everything in technology, it’s ultimately humans wielding this technology we have to fear.
PRISM, Upstream, and other monitoring programs employed by government agencies like the NSA take in way more information than necessary. And then there’s Google and Apple, Amazon, Facebook, Comcast, and everyone else who wants a piece of your data.
An organization like the NSA may say it exercises caution when wielding the power of device monitoring, but does it really?
As a whole, I’m sure Homeland Security, law enforcement, and everyone in the government wants to do the right thing. But there are individual bad actors in every bunch. What checks and balances truly keep these people from abusing the data we have available?
Quantum computing and AI may very well be our only defenses against other humans who may try to monitor us.
Increasing the Rat Race Pace
Speed is always the deciding factor in computing and gaming. Whichever gamer has the best gear will have the fastest reaction times in an environment where every millisecond counts.
The IBM Q is the first quantum computer on the market so far, and you can bet governments around the world are upgrading their supercomputers to have the fastest computers possible.
It’s all about how fast encryption can be broken, threats identified neutralized, and computations can be made to keep going.
Encryption is all mathematically based, and it’s the only thing keeping your personal information (bank accounts, SSN, etc) private. Without encryption, we’re basically naked to any threat – it’s like the door locks of our data.
And hackers don’t need quantum computers when they can place malware on computers to create a botnet. With cloud computing, the power of thousands, sometimes millions of devices can be used to destroy network servers.
This is why companies like Vectra are harnessing the power of AI to keep our computers safe from advanced threats. AI and machine learning can be trained to protect us from most threats, even AI-powered ones.
Par for the Course
With everything said, it doesn’t much matter if robots or humans rule us. We will inevitably be ruled, so the identity of our ruler is moot.
We depend on databases and computers to connect to each other and track our lives in every way. From self monitoring to government monitoring, we have no way of proving our progress and history without computer data.
You can’t take people’s word for fact these days – trust me on this.
So if you can’t trust anyone but computers, wouldn’t it make sense if they were in charge?
I’d argue they’ve long been in charge, and we’re just here to witness their power. So don’t worry about the rise of AI. Robots aren’t taking your job unless it was a menial, repetitive job you hated anyway. Dream bigger.