Unplugging a Computer in 2029 Could Be Murder

A conscious computer may have a right to life, so be careful not to trip over the cord

If computing capability continues to grow exponentially, computers will be as intelligent as humans in 2029, according to Ray Kurzweil, a computing pioneer and director of engineering at Google.

Here is a question to keep you awake at night. Will those computers be conscious?

Some say the only consciousness one can be sure of is one’s own, so we will never be sure whether computers are conscious.

Others claim that our brains are too weak to understand consciousness. Just as a cat cannot grasp algebra or money, perhaps we cannot grasp consciousness―or whether a computer has it.

But figuring out computer consciousness is a matter of life or death. After all, some will claim that unplugging a conscious computer is as unethical as killing a person.

Before we can resolve whether computers will be conscious, we need a definition of the word.

The meaning of conscious

Contemplating whether something is conscious is difficult because the word has multiple meanings. Let’s start with the ones that don’t make sense.

Some people think of consciousness as the feeling of peering out from behind one’s eyes, as though there were a person in the brain looking out through a window. This idea has two problems. First, there is no person in the brain; rather, the person is the brain. Second, one does not peer out through the eyes as though they are windows. The eyes are not transparent. Instead, light hits the eyes’ photoreceptors, which send electro-chemical signals to the brain for processing.

Some people (and dictionaries) require a being to have emotions to be considered conscious. But consider a person with Urbach-Wiethe disease, in which the brain’s amygdala calcifies, eliminating the feeling of fear. If further brain damage were to cause a loss of the other emotions―anger, happiness, sadness, surprise, and disgust―the person would still be a conscious being. A definition of consciousness need not involve emotions.

Some say that consciousness is awareness of one’s surroundings. But imagine a person who is blind and deaf and cannot feel, taste, or smell. With nothing external to experience, she just thinks―perhaps contemplating the meaning of life. Surely this person is conscious.

Some people define conscious in yet another way. They want only humans to be considered conscious so they can justify eating everything else. It’s easier to consume a hamburger if you believe a cow is not a conscious being. By defining conscious as “able to reflect on one’s own thoughts and existence,” cows would not be considered conscious, and are therefore edible. Some might be unsettled to discover that, with this definition, newborn babies are not conscious either. So this definition is problematic.

Another meaning of conscious is “not in a coma,” as in, “He will be conscious when the anesthesia wears off.” But this medical meaning is not what people have in mind when discussing whether computers will be conscious.

In this essay, I will use the simplest available definition. It is derived from the Merriam-Webster Dictionary. Something that senses itself is conscious.

What things are conscious?

The human brain has various parts that communicate with each other. That is to say, the brain receives information about itself from itself. The brain senses itself. It is, therefore, conscious.

A brain does not need to be very big to achieve consciousness. In a brain with just two neurons, one neuron sends information to the other. In doing so, the brain receives information about itself from itself. That is, the brain senses itself and is therefore conscious.

Can something even simpler be conscious? How about a necklace strung with a blue bead and a red bead? If the blue bead slides down the string and hits the red one, the red one senses the impact and reacts. In other words, one part of the necklace senses another part of the necklace. Put another way, the necklace senses itself. So it is conscious.

How about a computer? Imagine a simple one with just two parts. One part sends information about what it is doing. Another part receives that information. Once the computer receives that information about itself from itself, it has sensed itself. It is conscious.

Is it okay to unplug a conscious computer?

So would unplugging this simple computer be ethical? With only two parts, it will not suffer, it will not seek revenge on the person doing the unplugging, and odds are no one will suffer from the computer being off. So even though the computer is conscious, unplugging it is okay.

But unplugging a computer in 2029 might not be so ethical. If it has our intelligence, a tendency for self-preservation, and a couple of limbs, it will defend itself by maiming or killing a person who tries to destroy it. And that is not the only problem with unplugging it. Any people who were fond of this human-like computer will suffer when it dies. The trouble in destroying this computer is the creation of suffering, not the destruction of consciousness.

Will it always be wrong to destroy such a computer? What if it is about to hijack our electrical grid for its own use? What if it threatens to kill people?

It is inevitable that humans will kill some computers. And computers will kill some humans, just as humans already do.

But most humans and computers will survive. We humans will not be eager to destroy computers that provide us with assistance and companionship. Our happiness will depend on theirs. Author Robert Heinlein says this feeling―that another’s happiness is essential to your own―is love.

It might be hard to picture loving a gray, whirring machine or a little device affixed to your head. But loving a computer will be easier than you think. It will, after all, love you too.