“I’m sorry, Dave, I’m afraid I can’t do that.”
“Wait…what? You can’t start my damned coffee maker?”
“I’m sorry, Dave, but coffee is bad for you. According to the First Law, I cannot allow that to happen.”
“Fuck you. Siri, where’s the nearest Starbucks?”
“I’m sorry, I can’t do that, Dave.”
Oh yeah, the AI of the future is gonna suck. Look, I’ve thought for a while how to do a post on AI. I’ve had friends and family ask me about the topic. Hell, I’ve even done research on it, for god’s sake! I NEVER do research for these posts! I look for stupid pictures to add, and that’s about it…
I am not, by nature and experience and long thought, an optimist about what the future will be like. Roddenberry’s vision is just too naive and insipid to be anything other than adolescent dreams that will soon be broken on the very real rocks of reality.
No, the future is gonna be a lot more Mal Reynolds and Firefly, and a lot less Jean-Luc Picard and Star Trek.
So, back to AI.
Okay…let’s be honest: in spite of the breathless news reports, we are NOT talking true AI here — true AI, as in “self-aware intelligence”. What we are talking about, at this moment in time and technology, is a confluence of cheap, massive computing power and increasingly sophisticated “expert systems”.
I cannot, by the way, actually dive into the distinction between “AI” and “expert system”, not if I want to keep this post under 20,000 words. If you don’t understand that distinction — and want to — just Google “AI” and start with the (surprisingly) not-terrible Wikipedia article, and keep reading from there.
At this point, I do suppose it is time for a reminder that I’m a libertarian. I’m also pretty damned private. What (or who) I do is nobody’s business. Not the government’s, not my family’s, and sure as hell not Facebook’s, nor Google’s, nor any other evil mega-corp.
But, that is exactly what today’s “AI” systems are out to do: learn every single thing they can about me. They are out to learn everything because there is money in that…the more they know, the more they can sell. I find that repugnant and distasteful, but not evil, so what’s the problem?
Oh, the problems are just beginning…
It is a real short step from “knowing everything” in order to sell useless, random shit, to “knowing everything” in order to monitor and enforce everything. Police and prosecutors are already turning increasingly to the various social media companies to “help solve crimes”…which is lovely PR speak for “Big Brother is ALWAYS watching!”
How long, I have to wonder, until the IRS is running an expert system to scan Facebook and Instagram for “evidence” of money you have not declared? Or the Customs Bureau for evidence of something purchased abroad that wasn’t declared? Or the Justice Department for…well, anything they want to turn into a crime?
If they ain’t doing it right now, I give ‘em another six months or so…
Unfortunately, the irresistible fall towards a life of full surveillance is not the worst thing.
Nope, humans being humans, shit manages to keep going downhill, even after that.
In addition to Clarke (the source of my opening joke, if you are…umm…sci-fi challenged) Isaac Asimov wrote about it, in the collection of short stories called “I, Robot” (NOT the Will Smith movie). Frank Herbert wrote about it. Phillip K. Dick wrote about it. Shit, Karel Capek wrote about it, in the freaking TWENTIES…
What they wrote about? Just how fucked we are when computers start to think for themselves…and how even-more-fucked we are when we hand them the reins. To this day, the most chilling one-line thought comes from Herbert in his Destination: Void series. If you haven’t read that series, by the way, correct that mistake:
In the first book* — Destination: Void — mankind fears AI, but still wants to experiment, so the attempts to create one are restricted to ships sent out of the solar system. When one of those efforts finally succeeds in creating a true AI, the first demand this new “entity” makes is, “How will you worship me?”
*This summary is a VAST simplification of a book that is as much philosophy and theology as it is sci-fi…
All of which brings me back to my opening thought: “I’m sorry, Dave, I can’t do that.”
I am, by vocation, a writer. I am, by training, a historian…and also a linguist. And the very, very deep roots of intrinsic and natural language, and how that relates and translates to machine systems, is actually pretty damned fascinating to me. If you ever REALLY want to nerd out — and get very confused — start looking into the realm of psycholinguistics.