Let's pause for a moment and think about the term "artificial general intelligence."
Your house becomes self-aware.
The dishwasher doesn't just start automatically and sense which wash cycle to use, it knows how to fix itself on the fly and knows if you have washed a plate too many times.
Your car can drive on its own to pick up the kids, and it even has a dispenser for the snacks they like. They never complain, because the car is always on time.
More important, the self-driving car and the dishwasher are connected--your AI knows how much power each one uses and can effortlessly control them because it sees everything as a network endpoint.
Your brain is enhanced by the network. In fact, your brain gets better over time. So does your car. And your dishwasher. And everything in between. An all-controlling AI, known as artificial general intelligence (or AGI), is constantly evolving. Every endpoint is another opportunity to expand and grow and evolve, connecting everything and everyone.
That's the idea behind a company called OpenAI, which just announced a partnership with Microsoft. The most famous company in tech will invest $1B to create artificial general intelligence, something akin to what you see in the Terminator movies.
Now, AI experts hate to talk about Terminator. Yet the idea behind Skynet is perfectly relevant. An all-knowing AI would become self-aware. The part about the robot overlords is a bit far-fetched, sure. The danger, though, is real.
He knows too much about how dependent we are on AI, and that the AI routines embedded deep within computers are difficult for humans to track effectively. We have a hard time with the code that "sees" a semi pulling in front of a self-driving car. And Alexa doesn't always understand us. Bots are deeply flawed today.
The problem is that no one knows what "artificial general intelligence" really means. Some have predicted this godlike AGI won't become a reality for at least another century. It's not far-fetched to say an AGI will be the most impressive piece of technology ever invented, and the most dangerous. It is also not far-fetched to say the biggest company in tech already knows this, and is willing to invest massive amounts of cash.
We're not talking about individual feats and accomplishments, but an AI collective that can predict behavior, control every gadget you own, parse data related to city infrastructure, track the flow of traffic, analyze bridges and roads, answer your questions, and do anything else you can think of that might be helpful--correlating all of this data and constantly improving what it finds and what it tells you in the name of making your life simpler, more efficient, and less frustrating.
At the same time, the AGI can analyze itself. It is designed to improve its own capabilities, rewriting its own code to become more efficient and more powerful. It will be more than self-aware. It will be self-improving. The AGI, now properly funded, will aim at a super intelligence unlike anything we have ever seen before.
As you can guess, this will be both wonderful and deeply troubling. Elon Musk is right about the potential and the danger.
The question is whether we will listen on that second part.