My robot tried to kill me

Published 3:00 am Friday, September 27, 2019

When I was a child, my fascination for mechanical devices took root at an early age. My father always seemed to be working on something. He was always working on a vehicle, a piece of equipment, something around the house: he was rarely without a collection of tools and something under repair. My apprenticeship, of sorts, began with the tools.

I learned the names of the various wrenches and other objects. My father would ask for one, I would hand it to him. Over time, my level of interaction increased. Until one day, while we were conducting a bit of Southern delight, shade tree mechanic work, my father told me to handle a repair while he went to the auto parts store to collect a part for another project.

I was 12 years old. I meticulously disassembled the device, removed the broken part and started the tedious process of tracking my steps in reverse. When my father returned, I was finished and he tested the repair. It worked.

Sign up for our daily email newsletter

Get the latest news sent to your inbox

My hands were cut, I was bleeding, but the satisfaction of having tackled the problem on my own was incredibly satisfying. Now, 30-plus years later, I know that my father was just as proud as I on that day – I’ve witnessed the same successes in my own home.

In middle school, I dove into science fairs.

At one of the events, I presented a “robotic” hand. I worked on the device for nearly a year. It was a wild explosion of wires and tiny camera motors. The extent of the device’s capabilities was rather simplistic, by today’s standards. The hand could twist 200 degrees, flex and contract a few rudimentary fingers and pickup an egg. My robot was connected to my Radio Shack Color Computer via a serial cable. I wrote a simple program that controlled the power to the motors, allowing me to use the enormous computer as a basic remote control.

I was worried that my robot wouldn’t impress.

Not only did the device work, but it caused a minor amount of interest.

Fast forward a few decades and my fascination with robots hasn’t waned, but I’m cautiously optimistic.

I continue to tinker. I have robots.

A few weeks ago, I observed that one of my robots was missing.

The robotic vacuum cleaner wasn’t charging in its base. I didn’t hear the device cruising about the house.

I began my first search for a missing robot.

The associated vacuum app revealed that the device was charged, located in a back room of the house. When I went into the room, I didn’t hear or see it.

One of my children accompanied me on our 21st century search. Using a smart phone, a cellular signal, satellite GPS and our eyes and ears, we experienced difficulty locating the vacuum cleaner – the vacuum cleaner that wanders around our house, cleaning without receiving any appreciation. Now that it was “lost”, we were concerned.

The device was lodged under a recliner, stuck on a wooden leg.

As my son pulled the device from its captivity, we could hear the spinning of its wheels, the whirl of the vacuum. It had returned to “life”.

And then, something odd happened.

The autonomous robot surged forward, accelerated towards me. The robot struck my foot, ran up my shin and knocked me over.

The rise of the intelligent machines began in a bedroom in my house.

My son panicked. The robot detected my fallen body as an obstacle, avoided hitting my face and turned toward my son.

As you might imagine, he ran out of the room.

The vacuum, well – the vacuum started vacuuming.

I lifted myself from the floor and wondered if I should laugh or worry.

As a child, working on appliances, large machines and vehicles with my father was fun. Despite cutting my hands, the machines never cut me – I always hurt myself through carelessness or lack of skill. My truck has never chased me, my lawnmower has never bolted from the shed independently.

So, what did I do?

Obviously, the robot didn’t intend to knock me over. It was stuck. Its goal was singular: clean.  Once the disk-shaped device became wedged under the chair, it simply continued trying to clean. When my son dislodged the vacuum, its single-minded focus continued. The robot continued the cleaning imperative; I just happened to be in the way.

The device continues to clean the house, though I have a newfound and guarded skepticism about it.

When I read of concerns related to artificial intelligence, increasingly-complex robots, thinking machines, I’m encouraged. However, given that people are designing and building these devices, I have fears.

We’re not perfect, and our creations reveal our imperfections.

What if the single-goal-oriented device had been something else – let’s say the oven. If I were to speak aloud and command my oven to heat to four-hundred degrees Fahrenheit, I’d suspect the “robot” would do just that. However, what if the gas were leaking around the burning elements? Would the device be properly-equipped with intelligence to ignore the command and report a problem? Perhaps, in fact, I’d argue that the device shouldn’t be allowed in homes without adequate safety measures – routines that are designed to thwart the single goal of the appliance.

Should we worry about robots’ artificial intelligence?

Of course.

Should we fear it?

Well, I suppose that depends on one’s experiences and perspective. If your vacuum cleaner tried to kill you, you might be a bit reserved and reticent to embrace the winds of change.