When you are up to your neck in hype about AI, seeing an article that touts an electric toothbrush that is powered by AI makes you want to cry. Admittedly, the article’s angle is just south of cynical and tends towards the ‘nothing is sacred’ position. But still.
We are now getting into the Internet of Silly Things (IoST) territory (you really should click that link).
Such things divert us from the understanding that AI is a serious matter. Communities, companies and countries are piling into it, and beyond simply using it as a marketing gimmick. Mister Putin has said that whoever wins the AI race will rule the world. Mr Trump has mumbled and tweeted something or other.
AI is being heralded as the answer to countries’ military problems. He who perfects AI will be able to strike with lethal, laser-like efficiency.
Except of course, AI can be hacked and tricked, just like any other technology.
Tesla cars are a favourite target for hackers, possibly because it is fun taking the mickey out of Elon Musk, possibly because AI in a Tesla is, according to Musk, the most advanced. The serious point is that it is possible to hack a car’s system and make the AI believe that road markings are not where they actually are, that it is dark when it isn’t and a number of other smart little tricks, some with potentially lethal consequences.
At a more serious level still the AI in a missile could be made to believe that it was under attack or heading in the wrong direction. Defence systems could be fooled into thinking an attack was underway and react accordingly.
AI is very scary and very foolish at the same time. In order to keep ourselves sane, it is important to a) come up with an acronym for silly uses of AI (Bullshit AI Tales – BAIT?), b) list them and c) ignore the fact that AI could take us to the next level of humanity or kill us without really meaning to.