AI is an emerging technology that will drive the Next Economy. There isn’t a day that goes by when articles appear that predict or try to show how smart computers are the downfall of society. Most of these articles are heavily biased and ignorant, because they don’t truly understand what progress is being made by those of us working in the trenches.
One way we’ve been training computers to be smart is through video games. The most famous example is Deep Mind’s Alpha Go, which beat the best Go player in the world a few years ago. The news of a computer beating a human at a game caused a frenzy, not just in the tech community, across the world. People saw this as a sign that a future where computers replace humans is near.
But the proof is in the pudding.
As it stands right now, productivity is for the robots. Computers are really good at repetitive tasks that have a very specific goal; they suck at replicating how humans make their way through the world when uncertainty exists. This is one reasons why we’ve got a long way to go until computers are smarter and better than us at everything; not just video games.
A recent article from The Atlantic explains:
One characteristic shared by many games, chess and Go included, is that players can see all the pieces on both sides at all times. Each player always has what’s termed “perfect information” about the state of the game. However devilishly complex the game gets, all you need to do is think forward from the current situation.
Plenty of real situations aren’t like that. Imagine asking a computer to diagnose an illness or conduct a business negotiation. “Most real-world strategic interactions involve hidden information,” says Noam Brown, a doctoral student in computer science at Carnegie Mellon University. “I feel like that’s been neglected by the majority of the AI community.”
The bottom line: Everyday life is not like a video game. Real life requires making decisions based on limited information, something humans are very good at.
So if you’re worried that the end is near, don’t worry; it’s not. As with everything else, computing intelligence will evolve. Right now it’s very specific types of tasks that can be automated.