Artificial intelligence (AI) is the latest tech buzzword doing the rounds. Hardly a TV, car, or game can come out without the word being thrown around. As if just having AI in some way related to your product will make it somehow better. As with all technology trends, there is a fair amount of misunderstanding involved.
Evil thinking computers are so embedded in popular culture that the notion just can’t be shook. The truth is, however, that AI is not even remotely as dangerous as some believe. Not only is the technology not at any risk of turning sentient, the idea of AI sentience itself is nothing short of ridiculous. Let’s take a look.
AI Is Good At Everything
AI is some sort of super intelligent entity, right? Capable of doing any job better than a human. The notion that AI is good at everything is so wrong in so many ways. AI software is good at doing the specific task it was programmed for. The level at which it can do the given task is based entirely around the skill of the software engineer that wrote it. New AI software will have to be written for each new task.
To put it another way, thinking that self-driving AI would be good at voice recognition, for example, is the equivalent of dropping a fish on the beach and asking it to make a sandcastle.
AI Learns By Itself
While some AI is capable of adjusting on the fly, the software certainly cannot learn by itself. It is more accurate to say that programmers can write software able to make automatic adjustments, based on specific input. Again, let’s see that fish learn how to breathe air. In short ‘self-learning AI’ is another popular buzzword, and by extension has been represented extremely badly in popular culture.
AI Is Smarter Than A Human
There are so many levels to this misconception that just bringing it up can cause arguments. The best way to address the misconception is like this; calculators are better at you than maths, so why don’t you ask them to play some online blackjack real money?
The point is that certain AI is technically better than humans at certain tasks, but a long way from being smarter in literally every other regard. It is, essentially, like comparing the intelligence of a calculator to the average human. The comparison itself doesn’t make any sense.
AI Is Going To Become Sentient
The most common misconception is that AI will someday spontaneously become sentient. It is again a massive misunderstanding of what the technology is to assume it could ever stop being software, and start thinking like a human. This is the equivalent of suggesting that a Rube Goldberg machine could start talking.
It is, again, thanks to popular culture that the idea of AI sentience exists at all. But to say that we are a long way off from anything like 2001: A Space Odyssey is the understatement of the century. The 2001 in the title really was a highly optimistic prediction.
