AI ain’t so Smart. It’s Still a Toddler

robot vacuum

Artificial Intelligence has a long way to go before I get worried. It’s the big scare word these days. But if my experience with Google maps and my robot vacuum is any indicator, AI ain’t so smart. It is years away from being a real threat to humanity. Unless you consider it a threat because of its immaturity at this point.

AI is like a toddler

In human terms, I could say that AI is in its toddlerhood. I am surrounded by toddlers. What do an 85-year-old, a robot vacuum, and a 16-year-old have in common with a three-year-old? They all need supervision. At least the three-year-old and 16-year-old are trainable. Though that won’t show for years.

Even though Google maps and my robot vacuum are supposed to increase in their understanding of what I want from them as I continue to use them, which I do, they don’t really seem to get it and they often tend to regress.

Google Maps gets lost

For instance, every time I pull out of the driveway Google tells me to head north. I always head south because of the logistics of getting to the stoplight on the next street with my big truck. But after years of this, Googly Sue still does not get it and still tells me to go north. Halfway down the street she will reroute and usually tell me to go to a street about a quarter-mile away. Instead, I opt to make a left on the next street so that I can go to the light and get to the freeway. Every. Single. Time.

Robot vac gets stuck and falls down the stairs

My robot vacuum is wonderful in so many ways but it is not terribly intelligent. It has sensors that help it to avoid walls and other obstacles and it is supposed to not fall down the steps – I only have one – but it still does. Fall off the step. It also vacuums itself into corners and gets stuck under edges of furniture repeatedly. And even in the same session it will go back to the exact same spot and do the exact same thing. And I will have to rescue it. I learned early on that if I want this vacuum to really be of service to me and to last, I need to supervise it and help it out of situations that it gets itself into. Armed with this knowledge, I tend to have a very good experience with this vacuum. Which saves me much time and backache.

Even after training Google and my robot vac for months, if I do not use them for more than a few days, they will revert to their untrained state. Google will pretend to not recognize my voice and I will have to speak very slowly. And loudly. And I certainly cannot expect to use it in a hands-free mode. But that happens less and less as I use it more and more for voice typing. Which I do daily to save the strain on my fingers, hands, and wrists. And to jot down notes before I forget them, for my books.

Would you take a nap in a self-driving car?

So what brought this dissertation on was a discussion Dad and I were having about self-driving cars. He said, “I don’t want to get in the driver’s seat and take a nap.” I said, “But Dad, you could drive again. We could just program the car to take you where you wanted to go. All you would have to do is push a button and sit back and relax.” He wasn’t going for it. And I have to agree that even with self-driving cars, at least at this point, the human driver must absolutely stay alert and be ready at any moment to take over the controls.

Imagine the pilot of an airplane putting the plane on autopilot and taking a nap. Still expecting to get all the passengers to their destination safely. In the same way that a pilot must stay alert and monitor the controls of the plane, drivers of self-driving cars will have to do the same. If and when the bugs are ever worked out, there is still the problem of the system going down. Then the driver needs to take manual control in order to safely end up at their destination.

So although I enjoy the benefits that AI is bringing us, I am not going to worry about it taking over my life. At least not until it can clean my house properly.

Please share and like us:

Related Post