“The Vessel might help us to better understand the nature and origins of consciousness”

Q&A with Dr. David Harris Smith

Dr. David Harris Smith (McMaster University) is hoping that by attempting to build a conscious machine, we might be one step closer to solving a mystery that has intrigued both philosophers and scientists for generations: What is human consciousness and how did it come to exist in the first place? Smith says that their project will allow them to peer under the hood of the Vessel. Those insights may take the team into unchartered territory, exposing the processes that give rise to our own consciousness.

Do we know why consciousness originated in the first place?

This is a question about which we can speculate based upon theories that make sense from an evolutionary point of view. Why do we have what we call conscious experience? And what function might it be performing for us?

I think that the most plausible theory – and this is not my theory, this is the work of Princeton neuroscientist Michael Graziano – is that organisms depend upon prioritizing responses to information that’s coming in. Some information in the environment is more important to the organism than other information. So it’s the idea that consciousness evolved as a way of controlling attention, or of directing attention to the most appropriate sensory inputs.

I think if we are seeking to understand consciousness, it is very important to pay attention to the neuroscience. Neuroscientists will also attempt to verify their theories or their explanations by looking at cases where consciousness functions are unusual or fail, like when being unconscious. We also know of cognitive disorders, where certain components, that are normally associated with a healthy, conscious state, are not working properly. Some people, for example, suffer from what’s called depersonalization disorder, meaning they don’t have a sense of self. I think these are all clues that can help us to better understand the nature and origins of consciousness.

Do we understand consciousness well enough to build it?

I think, at this point, we understand some of the functions that consciousness must perform in order to qualify as what we call conscious experience. It has to feel integrated, it has to contain information about the present, about the self and the world etc.

But do we understand it well enough to build it? I would say we will always be able to understand it better. Maybe the question should be: Is there something to be learned about consciousness through attempting to build it? I would say yes, absolutely!

Do you think we will ever be able to develop a conscious machine? Or a machine that claims to have consciousness, like we humans do?

It’s easy to build a machine that claims to have consciousness. But we also need to understand what leads the machine to come to this conclusion. Just like us, the machine can only report on information it has access to. We, as humans, actually don’t have access to how our brains give rise to the experiences we have. We only experience the result, not how the experiences are constructed. Like, I claim to hear your words, but I actually don’t “hear” them. What’s happening, of course, is that sound waves or vibrations travel into the inner ear, sending electrical signals to the auditory centers of the brain where the electrical impulses are perceived as sound.

So here it becomes interesting: Unlike with humans, we’re able to crack the machine open. We can compartmentalize a part of the machine that is like that part of us that will naively report on its own experiences. The machine won’t know how it has these feelings, but it will know that it has a feeling of experiencing representations of the world and itself. So when we build a conscious machine, we will be making a model of what we think is happening, of the processes that we believe give rise to conscious experience. And unlike with humans, it won’t be a mystery to us because we will have designed it that way.

Could you tell us a little bit about how you are building the Vessel’s artificial consciousness? Is there a road map you follow?

Well, we’re learning as we go. What we’ve done so far is really attempting to understand the literature that’s out there and the research that’s been done. We are especially looking at the different theories about consciousness and how they support each other. We’re talking to as many experts as we can. This is also one of the reasons why the project is called the Vessel. We will always be able to add a new plank. It will carry on.

But do we have a road map? No, not really. This is really an open question. That’s where a lot of experimentation will be necessary, and a lot of consultation. What we’re doing right now is attempting to build a very, very simple visual system with a simple body that’s able to control and to explore its visual inputs and that’s capable of learning, almost like a baby.

How will the Vessel interact with humans?

All of us involved in this project are very curious about the human experience. We’re humanists, artists and architects and we’re very much motivated to engage with and understand what it is to have human experience – not just that, but also the quality of our experience. 

And I think one of our objectives in this project is the attempt to share that fascination about this profound experience of being present to ourselves and present in a world with others. 

So the interaction design of this artistic work will likely involve all the media art tools that we have at our disposal to resonate with the audiences, like storytelling, light, sound interactivity and much more.

AI technology has been advancing at a rapid pace, taking over more and more jobs in all kinds of sectors. Will humans be obsolete soon?

This project represents a very small, marginal sector of the entire field of AI and robotics. It’s mainly artistic and humanistic. And when we’re attempting to make human-like AI, in a way we’re keeping technology close to us. We’re attempting to include it in our social circle. 

And I think it is important that we do this because in some way we’re using technology to learn about ourselves, rather than to replace ourselves.