#trust doesn't rust

When it boils down to it, Karr and Kitt are autonomous vehicles. They are computers programmed to continuously make decisions to adapt to changing conditions. For example, in the first episode of Knight Rider, we see Kitt decide to speed up and pass a truck using pursuit mode rather than slowing down and staying behind it. This decision to speed up and pass was harmless, even fun, and it gave us a great character moment for Kitt before he even utters his first word onscreen. 

The problem comes when an autonomous vehicle is asked to make a much harder decision than a simple pass on a clear day. Roadways are dangerous. Crashes happen. Even if a computer like Kitt is a perfect driver, other drivers will make mistakes. How should an autonomous car respond to an incoming accident? The obvious and easiest answer is to program the computer to “minimize risk”. But how does one go about this? Not to get all angsty on you, but if Kitt is forced to make the decision to hurt someone else to save Micheal’s life, would he do it? Should he do it?

The problem with having a computer answer an ethical question like the classic Trolley Problem (google it if you don’t know) is preprogrammed intent. See, a human makes a life-or-death decision based on reaction. If a human tries to avoid an automobile accident and makes a snap decision that ends up with another driver hurt, the human typically isn’t blamed- they made a split-second reaction to the best of their ability. Where this differs with a computer is that a computer defers to what it’s been programmed to respond with. If an accident happens, a computer will have already decided who it is going to hurt in order to “minimize risk” overall. The human didn’t intend to hurt anyone else. The computer full well intended to hurt someone else, even if to preserve other human lives instead. Best case scenario, that’s manslaughter. Worst case, murder.

Now, Knight Industries (and by extension the Foundation for Law and Government) actually handled Kitt rather well. Kitt gets around a lot of these decisions because he himself is nigh invulnerable. Because he (and therefore any human passengers he carries) are able to survive much more dangerous crashes than an average vehicle, it’s easy enough for Kitt to minimize harm done by choosing outcomes that will only damage himself. For example, in an accident scenario, he might choose to roll  off a cliff instead of getting into a head-on collision with another vehicle. 

Kitt also gets around a lot of these ethical dilemmas by having Micheal. By having a human driver to defer to, Kitt doesn’t have to make these sorts of decisions alone, and in times of true ethical crisis, he doesn’t have to make the decision at all. Turning control over to Micheal puts a human with human ethics back in control of the situation, releasing Kitt from any blame. 

So what does this all have to do with Karr?

I hate to say it, but from this angle, Karr is terrifying. Imagine a self-driving car like Kitt, confronting all of those same ethical dilemmas, but Karr has zero regard for human life. Karr, as presented in “Trust Doesn’t Rust”, will always choose the situation that is best for his own survival, regardless of how many humans he will injure or kill. Everything I just mentioned about the ethicality of a computer choosing to kill one human to save many? All of that goes out the window for Karr, because for him that’s not even a question. With his programming, the question of whether Karr would hurt anybody is not an “if”, it’s a “when”. While Karr didn’t hurt anybody (onscreen) during the span of “Trust Doesn’t Rust”, the longer he remained operational, the greater the odds of him encountering a situation where he would see killing a human as the best option for saving himself. It would be murder, no question about it.

So why the hell would Knight Industries ever, ever “forget” to program Karr to protect and uphold human life? What kind of idiots would skip this vital step?

Well. . . we actually have this same exact problem happening today, right now. The U.S. military is looking into building truly autonomous drones to conduct drone strikes against high-value targets. In this scenario, programming an armed drone to “minimize harm” to human life would be contradictory. Ideally, a drone strike against a dangerous target would save more lives in the long run, but if said dangerous target has surrounded themselves in innocent civilian, then should the drone still take the shot? Or should it let the target escape and hope for better luck next time?

I suspect that Knight Industries originally intended for Karr to work at a much higher level of law enforcement or perhaps even in the military. A highly-mobile, indestructible transport would be a boon for hunting down terrorists, right? And in order to fulfill this role, Karr would have to be able to make the decision to take human lives. The intended purpose was always for the greater good, but I think sometime through the project, Knight Industries realized that it would be unethical for a computer to be making those types of decisions, so they pulled the plug. How much of this was under the control of Wilton Knight is up for debate.

What I do blame Wilton Knight/Knight Industries for is abandoning Karr fully programmed and assembled. It might make sense to keep Karr’s body around as a prototype for developing Kitt, but Karr’s microprocessor should have been destroyed or at the very least reprogrammed and recycled once it was found to be ethically unsound. The fact that Knight Industries let this obligation get away from them is irresponsible and put everyone, (including Karr himself, as a sentient being,) at risk!

When Karr was reactivated, that alone was reason enough for Devon to order him destroyed. When Karr began displaying dangerous behavior, it became morally imperative for Devon and Micheal to stop Karr at all costs.

And therein lies the biggest tragedy of them all- was it really Karr’s fault? I think the answer to that is no, and that’s the worst part! It’s not like Karr can control how he was originally programmed, and it wouldn’t be morally responsible to let him stay online long enough to learn and grow like we’ve seen Kitt do. Karr would never consent to being restrained because of his trauma, and there was no feasible way to capture him against his wishes because of his capabilities. The only solution was his deactivation, and that’s tragic as hell.

That’s why I can agree that what happened to Karr was a tragedy, but still believe that Devon and Micheal made the right call to destroy him.

Am I reading into it way too much? Absolutely! Am I applying real-world concepts that the creators of Knight Rider probably never intended to address? Absolutely! Are you more than welcome to disagree with my analysis? Absolutely! I also am not judging or calling out anyone who likes him and thinks that what happened to him is unfair. I’m definitely not saying that he’s not worth redeeming or that if you redeem him, you’re a bad person.