So far in our series looking at the different types of Artificial Intelligence we have looked at Narrow and General AI. We’ve covered how Narrow AI is custom created to handle very specific tasks, while General is more akin to a human being. At this stage we now turn our attention to Super AI.
Super Artificial Intelligence
From the other two approaches it’s logical to place Super Artificial Intelligence as a step beyond the capabilities of any human being. And that’s exactly it; Super AI is better at everything a human can do. Super AI would be able to process paperwork, determine where life in the Universe may exist, play Scrabble, fly a plane or look after your pet better than you. It’s the AI most commonly found in science fiction, and the one that the public would most readily perceive AI to be.
You may be tempted to think Super AI already exists. After all, didn’t IBM build a supercomputer that can beat humans at chess? While that’s true, the same supercomputer couldn’t boil you an egg or discuss the meaning of life. It was built for one purpose and so it’s not better at anything than anyone alive (or dead).
In the same way, the digital assistants such as Google’s Alexa is far from being considered to be Super AI. These plugged in devices merely answer questions you ask based on data available from the web. They’re no more than a search engine with speech nailed on the front and back. And I know someone will tell me off for being a little crude with that description but it’s essentially correct.
So, when will Super Artificial Intelligence appear? The short answer: don’t hold your breath. Super AI is a long, long way off, further away than Artificial General Intelligence and extremely unlikely to appear in this century. So we’re going to see Super AI pop up in lots more science fiction movies before it finally becomes a reality.