AI Quandary

Sometimes I forget how long I’ve been writing this blog. I have been thinking about AI (Artificial Intelligence) and knew that I had written something about it before, but didn’t realize that it was twelve years ago! Ha! That post still has some merit yet today. (You can find it here.) It was about self-assembling robots. We’ve come so much farther than that now. Companies and Countries are racing to be the first to perform the AI equivalent of Catching Lightning in a Bottle, while seemingly mostly ignore that even in the best case scenarios, it is still Playing with Fire. Very smart and very wealthy people like Elon Musk and Bill Gates are throwing money at this endeavor while confessing the outcome may not necessarily be good.

I’ve been reading science fiction since I was in grade school and looking back at it, it can be eerily prescient. They have addressed artificial intelligence in a myriad of ways. The stories of it going wrong vastly outnumber the handful of positive outcomes. The general theme is that eventually the AI sees the flaws in its creator. From there, it either moves to take-over to protect and save us from our incompetence or decides we’re a threat that must be removed. In movie themes, the former is “I, Robot” and the latter is “The Terminator“. Both of those movies show the resilience of humans and our ability to fight back. In my opinion, the reality of that is questionable. In most cases, AI inherits our hubris, which might not save the puny humans, but doesn’t bode well for human created AI in the long term.

The problem is, there is no good way to put controls on this. Even if all the worlds governments were able to agree on controls, something like Isaac Asimov‘s Three Laws of Robotics, and could be trusted to follow them, there’s no guarantee that some Dr. Evil out there, working in his lair under his volcano island, doesn’t figure it out and unleash it on the world. (Not to mention that more than a few Asimov stories were about loopholes in the Three Laws…) So now, much like the mutually assured destruction by atomic bombs from the Cold War, we’re in an AI arms race to make sure our AI can defend ourselves against their AI.

When The Terminator came out in 1984, Judgement Day, when the AI became self-aware and took over, was August 29, 1997. We’re well past that mark, but if you look at things we have today, such as the Internet of Things (IoT), it would be much easier today than it was projected back then. We’re all carrying around pocket computers that listen to us constantly and have made it convenient not to know anyone’s phone number or even how to get anyway, because it does all that for us. We all have Alexa or Google Assistant listening for our commands or talking to our other devices in our homes. Bringing all that information and processing power under potentially malevolent control should give us pause. Yet we go blithely on…

The whole TicTok controversy is the perfect example. I understand the U.S. Government’s concern that the information gathered could be shared with a foreign adversary, but the real problem is that the data gathering is happening at all! Regardless of who has access. No one, and particularly young people, spend no time reading the privacy policies for the apps on their phones. Nor do they seem to care. Why does the cool flashlight app need access to my contacts and location? Oh, well… I want the app, so just click “Okay”… We all know Google and Facebook track us, but they’re soooo convenient and fun and all our friends use them…

All of this information is out of our control once it hits the internet and is stored in the Cloud. The European Union has instigated privacy laws requiring companies to delete private data upon request, but how can that be confirmed? And the U.S. and other companies don’t have those requirements, so data storage here is unlikely to be protected.

We can’t all become Ted Kaczynski and move to the woods. His bombing campaign was ineffective in changing the course of technology in the 1970s anyway. But we do need to be aware and wary. AI is just the latest technological quandary we face. There will be more…