• Contribute
• What's New
• Submit News

Generation5 already had the opportunity to interview Steve Grand. Now a mere two years on, Steve is recognized as one of the leading programmers and engineers in the Artificial Life field. Generation5 is incredibly fortunate to be able to interview Steve again. Please visit his new webpage here.

G5: You have an incredibly interesting concept of life; atoms and human beings are not things, they are patterns that have learned how to survive. This surely has huge implications when applied to computing and ALife. But surely the idea also extends outwards -- is the society intelligent? Indeed, is the world or universe intelligent?

Absolutely! I think it is very probable that societies are intelligent, to a degree. Maybe the entire universe is, in which case some people might choose to call the result "God". James Lovelock got a lot of criticism for his theory (Gaia) that our planet, complete with all its life forms, is a self-maintaining (and therefore somewhat intelligent) pattern. Yet it seems pretty self-evident to me. The snag, of course, is that it is extremely hard for us mere organisms to imagine what it would be like to be a society or a planet or any other self-maintaining pattern larger and more diffuse than ourselves. Presumably, a neuron or a liver cell (if it were able to think about it) would find it equally hard to see how it and its trillion other loosely coupled compatriots could possibly interact to constitute a single, intelligent being.

That's a very good question. I really have no idea what consciousness is. People often use the word in different ways, and sometimes use debased definitions that even a thermostat would satisfy (since it is "aware" of its environment). I think the most relevant question about consciousness is not so much what it means to be aware or even self-aware, but what exactly is it that is being aware. It sure as hell isn't our bodies as physical entities, though I don't imply anything metaphysical by saying that. The only thought I can offer so far is that the being inside of my body that is conscious is not, I think, my brain itself, but "something" that exists inside and because of a virtual world created by my brain. In other words, in order to be conscious - to contain a conscious mind - our brains must have an imagination. We live in the virtual world of our imaginations and not in the physical world of our senses. Insects, to take one example, don't appear to have the neural circuitry that would allow them to support a non-trivial mental model of the world: they have no imagination and so are not conscious. I doubt if societies have the capacity to support mental models either (although they have some of the components, such as memory).

So yes, I think alive and conscious are different things. I don't think either is black and white - things are not simply alive or inanimate, conscious or unconscious, but there are many different kinds of life and consciousness.

The thing about norns that I believe qualifies them as living is that they are not the result of a computer program pretending to behave like a living thing; they are the result of a computer program pretending to behave like neurons, biochemicals and genes. These are not real neurons and chemicals; they're just pretend, but the thing we call a norn is something greater - an emergent consequence of the interactions between these sham neurons and chemicals. It's not the norn's fault that its components are a sham, and it is no less real because of it. The behaviour that norns show is not directly programmed in but emergent. It is also self-maintaining structure (norns adapt, metabolise and reproduce) and therefore in my view norns qualify as alive.

I wouldn't claim that norns are conscious, however, because they are trapped in the world of their senses and can't step aside from reality like we can. They can't have independent thoughts and indulge in flights of fancy.

Lucy is my big attempt to solve the next step in artificial life: how to build a neural substrate capable of exhibiting imagination - a neural network that can hypothesise, plan, anticipate, hold intentions and beliefs, and even dream. I'm trying, as well as I can, to bring Lucy up as a baby. In other words, I want to test out my theories by watching her learn the things that a baby has to learn: how does she move one arm without moving the other? Which bits of the world belong to her and which are independent? How does she move her mouth and throat to make sounds like those she hears coming out of that big thing looming over her (me)?... If I can get that far, then I'll build a new body for Lucy that's bigger and stronger and see if she can learn to crawl and then walk. Whether I can do this depends on whether my hunches about the brain are right, and that remains to be seen. But I'm exploring some new stuff - an active approach to perception, and a neural topology that's highly recurrent and dynamic. Something interesting will come out of it. Don't hold your breath waiting for complex natural language and logical reasoning, though - remember that even Albert Einstein took five years just to learn to tie his shoelaces!

We play god every time we conceive children (or risk conceiving children), and some people put very little thought into that decision! I think artificial life challenges us to confront our many unquestioned assumptions about life: when it begins, when it stops, how we should treat it. Hitherto, we've been trying to make ethical and moral judgements based on a case study of one (the form of life that evolved here on Earth). A-life helps expand this to a sample of two or more, and this must surely help us to understand life better. Imagine trying to make any general judgements about the concept of fruit, when all you've ever seen is a banana! If a few artificial lifeforms (none of whom are very much alive yet) have to die in the name of progress than that can't be helped. After all, we're killing ants and spiders and monkeys and gorillas and tomatoes and other human beings all the time, essentially through ignorance of whether it is right or not.

They seem to have different connotations, somehow. Everyone is fascinated by life - what it is and how it ticks, but they often draw the line at meddling with existing forms of life. Often this is for good reasons, but sometimes for poorly thought out ones. There are people who find the concept of artificial life scary too. This is at least partly due to certain of my scientific colleagues who spout utter drivel about how machines will grow rapidly more intelligent than humans and will take over the planet. Why their doom and gloom scenario is wrong is a rather involved topic, but in my view it is logically indefensible, irresponsible scare-mongering and says more about the screwed up lives of the "scientists" concerned than it does about artificial life. I hope I've done my bit to show that artificial lifeforms can be friendly, fascinating and accessible, and I'm certain that research in this area can only benefit us and our descendants.