People get so excited these days about technology. Everything is getting smaller and more integrated. Computers are faster and smarter. Some gleefully proclaim the day is coming when we’ll all have cybernetic implants that allow us to do anything. Nano probes roaming our blood stream will be ever on the lookout for problems to fix. The lactose-intolerant will be able to consume dairy without the unpleasant side-effects. Shop class table-saw amputations will become little more than an inconvenience while we wait for the severed bits to grow back (and don’t tell me teenage boys who know there are no permanent consequences won’t take FULL advantage of that!). Huffing glue will be perfectly safe (and become a national pastime) since we can just restore our memories and personalities from a backup on ‘the Cloud’ (I want to fart every time I hear that term) once those nano probes repair all the brain damage.
Doesn’t that sound great?
Even better, these Singularity fans crow, will be the dawning of Artificial Intelligence that accompanies the Singularity. AI inside our cybernetic implants will provide us with a range of useful tools, from being a limitless resource for cheating on exams, to being our BFF so we don’t need to find or interact with anyone else, to playing Devil’s Advocate to the other voices in our heads when we go off our medication.
While I’m as excited as the next guy by the prospect of one day having a piece of hardware that can sass me jammed into my body (the level of excitement varying depending on just where it’s jammed), I don’t want to focus on that. I want to talk about the AI component of this budding fiasco.
People are ignoring an important problem with the arrival of AI. I’m not talking about the machines rising up against us and either enslaving us or wiping us out (though I suppose that bears looking into as well). I’m talking about degrees. Degrees of intelligence.
Think back to when you were in grade school. Remember that teacher’s pet suck up who always had her hand up, always knew the answer? Annoying little twit, wasn’t she? And let’s not forget the other end of the spectrum, the class clown who couldn’t think his way out of a cellophane bag (nor had the intelligence to realize just how fatal sticking his head into that bag would be, rest his soul). What makes you think AI will be any different?
Sure, we’re gonna have state-of-the-art, MENSA-busting AI manufactured at the finest, highest-tech institutes. And those AIs are gonna be smarmy gits none of us can stand to be around, let alone play Jeopardy against. And then, within a few years, we’re gonna have the cheap knock-off AIs, manufactured overseas by outsourced techies who have no respect for intellectual property laws. The quality on those will be all over the map, but I can guarantee you that they’ll be cheap, which means they’ll be everywhere. If we’re lucky, they’ll speak English. The Department of Motor Vehicles, MediCare, Social Security, and any other government bureaucracy you can think of that makes use of call centers will be answering their phones with AIs made by the lowest bidder. How do you think that’s gonna work out for us?
Sometimes I think we’d be better off getting wiped out by the machines instead.
So what do you think? Am I a crazy nay-sayer and doom-prognosticator? Do you think the AIs will be smart enough to avoid / fix the problems I foresee? And for that matter, just how do you hard code the Three Laws of Robotics into a machine? What if there’s a bug in that code copied-and-pasted by a Chinese programmer from an old version of an open source Linux module with known vulnerabilities? Do you really think we’ll be safe?