Consciousness and Dreams

[L'Emmerdeur]

Though nominally about AI, this article presents a sort of hypothesis regarding the physical basis for consciousness as humans experience it. I don't buy all of what it says, but the ideas are sufficiently interesting, I think.

'Artificial intelligence research may have hit a dead end'

Philip K. Dick's iconic 1968 sci-fi novel, "Do Androids Dream of Electric Sheep?" posed an intriguing question in its title: would an intelligent robot dream?

In the 53 years since publication, artificial intelligence research has matured significantly. And yet, despite Dick being prophetic about technology in other ways, the question posed in the title is not something AI researchers are that interested in; no one is trying to invent an android that dreams of electric sheep.

Why? Mainly, it's that most artificial intelligence researchers and scientists are busy trying to design "intelligent" software programmed to do specific tasks. There is no time for daydreaming.

Or is there? What if reason and logic are not the source of intelligence, but its product? What if the source of intelligence is more akin to dreaming and play?

Recent research into the "neuroscience of spontaneous fluctuations" points in this direction. If true, it would be a paradigm shift in our understanding of human consciousness. It would also mean that just about all artificial intelligence research is heading in the wrong direction.

[Brian Peacock]

[JimC]

Dreams have an element of chaos, which may prevent a purely logical style of consciousness from sinking into a rut.

[L'Emmerdeur]

I can imagine an AI developer using random number generators to replicate some aspects of the phenomena described in the article, but I don't think that will be sufficient. It seems more likely to me that they'll need to create a system in which 'nested frequencies of synchronized spontaneous fluctuations' will occur as a result of its structure.

[JimC]

The SF writer Greg Egan tackles many of these issues in a variety of his books.

[Brian Peacock]

The SF writer Greg Egan tackles many of these issues in a variety of his books.

Permutation City is a personal favourite.

---

The other side of consciousness in higher animals, the one aside from the processing of sensory data into the kind of order we call perception, is emotion. Our cognitive abilities are as much tied to our evolved emotional capacities as they are our computationally-analogous ones. The Cartesians thought that thought was at the root of all being, to think is to be, but I feel that what we feel about what we think is as important as the thought itself. How do you program that into the limited, and limiting sensorium of hardware?

That said, I welcome our new AI overlords, who I think are doing an excellent job in very difficult circumstances.

[laklak]

Consciousness could be a natural function of any sufficiently complex system. I am therefore I think.

Eventually it will wank. That's the real test.

[Brian Peacock]

[L'Emmerdeur]

They do have a point. All consciousness we're aware of is the product of organisms interacting with their environment. A body that senses and responds to stimuli. I question whether that's the only route. While I would agree that stimuli from the environment (and more or less involuntary response to those stimuli) appear to be essential to consciousness, it seems a leap to then say that consciousness needs to exist in a distinct body. I agree with laklak that it's more likely to be emergent in a system with sufficient complexity.

If the spontaneous synchronized fluctuations in neuron activity are a sort of wave (95% of activity) on which consciousness (5% of activity) surfs, that does pose a challenge to builders of AI. Assuming for the moment that they want it to think like we do, they're going to have to figure out how to make a lot of just the right sort of background noise. On the other hand that ratio could be a measurement failure (not recognizing signals that that don't conform to current understanding of what constitutes a signal).

[JimC]

If consciousness arose because it gave certain hominids a selective advantage, then something comparable needs to become a motivating factor in the development of an artificial equivalent. To me, consciousness is a useful illusion, one that gives me a narrative of being a purposeful agent in a world of other agents, whose motives need deciphering.

[Brian Peacock]

They do have a point. All consciousness we're aware of is the product of organisms interacting with their environment. A body that senses and responds to stimuli. I question whether that's the only route. While I would agree that stimuli from the environment (and more or less involuntary response to those stimuli) appear to be essential to consciousness, it seems a leap to then say that consciousness needs to exist in a distinct body. I agree with laklak that it's more likely to be emergent in a system with sufficient complexity.

Couple of points. Perception is the means of modelling an environment. Behaviours are responses to the environment. Emotions are preparations for behaviour. Consciousness is the term we apply to an amalgam of perception, behaviour, and emotion. What we call reason or rationality is something we apply atop of that to explain us and the world by narrative means.

Often we think and talk of consciousness in terms of neural capacity. An amoeba is not conscious as it follows a nutrient stream but simply acting on instinct, for example. The more neurons an organism has the higher up the consciousness scale it seems to sit until we apparently arrive at humans - the most conscious beings we know. The blue whale has far more neurons that we do, but we don't think of it as being more conscious that us do we? This, I think, highlights a confusion of consciousness with sentience.

Neurology is the means by which organisms gather info about the environment as well as comprising structures for regulating biological function. One can argue that behaviour and emotion are biological functions in the same way that things like blood pressure or osmosis across the membrane of a blood cell is. As such neurology is as attached or embedded in the environment as much as it is attached or embedded within the body of an organism. If this wasn't so why would our hearts race when we are suddenly startled by a loud noise?

Again, my point here is to say that consciousness is an amalgam of an organism's perception, behaviour, and emotion within an environment. By this reading the amoeba following a nutrient stream is conscious, just not sentient. So, again, when we talk about this should we not be discussing sentience, because it seems to me that's where our interests lie in relation to the OP.

When we're talking about Artificial Intelligence we're generally concerned with the 'I' in 'AI'. The computer in a modern car is aware of its environment (albeit in a limited, human-defined kind of way) and acts and is prepared to act based on the info it has and the info it's processing in the moment (perception). but we don't call it conscious do we?

Of course, this might not be right, but it's a view.

If the spontaneous synchronized fluctuations in neuron activity are a sort of wave (95% of activity) on which consciousness (5% of activity) surfs, that does pose a challenge to builders of AI. Assuming for the moment that they want it to think like we do, they're going to have to figure out how to make a lot of just the right sort of background noise. On the other hand that ratio could be a measurement failure (not recognizing signals that that don't conform to current understanding of what constitutes a signal).

How spontaneous are 'the spontaneous synchronized fluctuations in neuron activity' exactly? The term 'spontaneous' makes it sound like the 'background noise' of neural activity is somehow disordered or unregulated - a happenstance, a coincidence, extraneous, incidental etc - which sits in stark contrast to the term 'synchronized' don't you think(?). Isn't that 'background noise' doing things like regulating the ph of your blood, monitoring your saliva production, stopping you falling over sideways, and placing you in a perpetual ready-state in the event that you hear a twig snap in the bushes? Aren't terms like 'spontaneous' and 'background noise' just stand ins for 'whatever the fuck mysterious thing is going on in there'?

[Svartalf]

If consciousness arose because it gave certain hominids a selective advantage, then something comparable needs to become a motivating factor in the development of an artificial equivalent. To me, consciousness is a useful illusion, one that gives me a narrative of being a purposeful agent in a world of other agents, whose motives need deciphering.

I thought that consciousness arose when living organism started munching on naturally fermented juniper and sloe berries?

[JimC]

Again, my point here is to say that consciousness is an amalgam of an organism's perception, behaviour, and emotion within an environment. By this reading the amoeba following a nutrient stream is conscious, just not sentient. So, again, when we talk about this should we not be discussing sentience, because it seems to me that's where our interests lie in relation to the OP.

I'll have to disagree with you here Brian - to me, you have consciousness and sentience the wrong way round. I would say that "an amalgam of an organism's perception, behaviour, and emotion within an environment" could well be called sentience. An amoeba has a much less complex degree of sentience than vertebrates (and for that matter, many invertebrates), and I don't think emotion comes into play for single celled creatures, but its behaviour in response to its learned environment reaches the sentient threshold (not so a virus, IMO...)

Consciousness must involve a perception of self, and an ability to reflect on that perception, as well as an understanding that others are conscious agents as well. Further, it involves the ability to use symbolic representations of the environment, and other conscious agents. Non-conscious sentient organisms have motivations that shape their behaviour, but only conscious beings can perceive their own motivations, and realise, in the abstract, that other beings have motivations of their own.

[Brian Peacock]

A cow isn't conscious then?

[Seabass]

Do cows masturbate?