Are Robots…Alive?

In thinking about what to write next, I went over to Google and typed: “Are robots”… and the magic of Google’s prediction engine responded with “alive” as the second completion. In case you’re wondering, the first was: “Are robots real?” (Yes, I know).

So what people wonder most about, regarding robots, is whether they are alive. That’s interesting. Not the answer to the question itself (which I’ll give you anyway in a moment) but the further question of what we are actually asking when we ask whether robots are alive.

What is the real question in the mind of the questioner who is asking this question?

It’s unlikely that so many people are interested in the intricacies of the scientific definitions of life. It’s more likely that this question is a kind of proxy for some other question, which the questioner didn’t know how to frame better. For example: Are robots conscious? Or: Should robots have rights like living beings do? Or: Do robots have an instinct of self-preservation? All good questions.

So, are they?

Let’s get that out of the way first. Living things have a few common properties:

1. Living things take energy from the environment and use it for their own self-organisation and growth.
2. Living things have some mechanism of procreation that allows them to create new organisms.
3. Living things are autopoietic, which is Greek and means: self-making. This means that they are able to create and uphold their own organisational structure, even in the face of disturbances. So, for example, if I cut a piece of paper, it will stay cut for ever. If I cut my finger, it will heal.
4. One could also add that living things tend to adapt to their environment. Depending on how one understands the word “adapt” here, this might or might not be the same as number 3 above.

The problem is that none of these are strictly necessary, and neither are they sufficient for life. For example, viruses don’t procreate without a living cell as a host. So, strictly speaking, they are not themselves “alive.” Some animals, like mules, are sterile, and would therefore not count as “alive.” Obviously, that won’t do. And regarding adaptation, do we mean adaptation of one organism to its immediate environment, or the adaptation of a whole species over time? These are very different processes.

Recently, researchers have proposed “lyfe” as a concept that is supposed to be more inclusive and that would apply to the stranger forms of living stuff that we may find on other planets. NASA reportedly uses as a working definition: “Life (lyfe?) is a self-sustaining chemical system capable of Darwinian evolution.”

If all this sounds terribly vague, it is because, well, it is. Nobody really knows what life is, any more than we know what other fundamental concepts like “art,” “beauty,” and “freedom” mean. We use them, we may even think that we’d “recognise them when we see them” (a once-infamous definition of pornography), but they resist precise definition.

Anyway, trying to apply these as best as we can, we must conclude that robots probably are not alive.

Photo by Scott Webb on Unsplash

Energy, information and growth

For one, robots do extract energy from the environment (for example, when they charge their batteries), but they don’t transform this energy into self-organisation and growth. Today’s robots are static devices, once designed and manufactured, and they don’t change or grow in any material way. Now one might try to argue that robots do grow in the sense that their software grows: they learn new things, their neural networks create new connections and so on. But these cannot be properly described as processes of “life.” These are processes of “learning.” As we would distinguish in our own bodies between, say, digestion and studying (a life function vs an information-processing learning function), so we would also not count the growth of neuronal synaptic connections in a robot brain as a life-sustaining function. In all material ways, robots do not (yet) grow.

But even this can be disputed. Autopoiesis (self-creation, mentioned earlier) is a term coined by Chilean philosopher Humberto Maturana and biologist Francisco Varela. They thought that actually every kind of interaction of an organism with its environment is the same as every other kind; so that digestion is not really anything different from reading a book. Yikes.

The thought goes like this: when I read a book, photons reflected from the surface of the paper reach my eyes and there a biochemical process leads to some neurons firing and, in the end, inside me, all these biochemical disturbances somehow convey “information” to me about what is written on that page.

But I can say the same about what is placed in front of me as food on a dish. Molecules of food leave the surface of the dish and reach my mouth, where a biochemical reaction with my tongue leads to neurons firing… and so on. At the end of that chain of events, I’ve also received information: about what’s on the dish. Seen like that, even eating is a process that gives me knowledge. Described like that, eating become almost indistinguishable from reading.

Things get really complex here, so we better stop and look at the other criteria.

Where are the kids?

The ability to make children, smaller copies of oneself, is another sign of life. Already here we should pause, because actually, according to this criterion, none of us is alive. We all need a second person to make a child. So if we believed that procreation is a necessary sign of life, then the basic living unit would not be “a human” but “a couple.” Everybody else, every single human being, would be something more like a virus, a rudimentary thing that has a part of the ability to live, but requires a suitable host in order to actually multiply.

Again, robots don’t do this, so they’re not alive. But, sadly, neither are we.

Adapting to the environment

We have both individually the ability to adapt to changes in our environments and as a species. Individually, we will notice that a chair in the living-room has moved, and not attempt to sit on the empty space where the chair stood yesterday. But a robot wouldn’t either. This level of adaptation is necessary for every artefact that will interact with the real world. All modern robots have it, to some extent, self-driving cars have it, even mobile phones that adjust the screen brightness to the ambient light have it. So that’s not a deal-breaker for robot life.

So, are robots alive? Yes; and no; and maybe. And maybe we ourselves are not. But those three species that we kill off every single hour… They certainly were.

Have a nice weekend!