Although I am not a professional scientist, I do have my day job in Science education. Science and all that it stands for means a lot to me. I think its slow arrival over previous centuries amounts to the most significant cultural change in human history. You might be asking yourself: why?
Last week I mentioned the subjectivity/objectivity issue in studying consciousness, which for many has been the main stumbling block when attempting a scientific description of the human mind, related to the issue of qualia and the “hard problem.” We are each the sole experiencers of our own mental activity. Direct contact with others is impossible. But because we have enormous quantities of indirect contact on a moment-by-moment basis, we all have the opportunity to make a model of other human beings in our minds, using ourselves as an exemplar, models which become more accurate the more empathic we are. Consciousness is empathy, to expand this loose metaphor, although even in human beings it can be selective.
Chimpanzees may have good theory of mind, but they are quite capable of killing and eating one another. Such acts are deeds more of instinct than anything else, where compassion, though present, remains for whatever reason unused. Human beings on the other hand are born with a powerful drive to empathy, and have to be taught to repress it in order to kill others. Military techniques of dehumanisation apply to both to the enemy and to the soldier, and are aimed at reducing as far as possible the use of empathy.
If the study of consciousness is to spring from a scientific foundation, it is probably best to first distinguish between the Scientific Method and Science. The former is the technique used by scientists to ensure that the descriptions of the world they work with are as accurate as possible – that is, that the description really does match reality. To do this, the Scientific Method begins from the standpoint of human beings deferring to reality. It does not work the other way around. Hypotheses can then be experimentally tested, honed, altered, and even thrown away to make way for a better hypothesis, before, following further testing and peer review, hypothesis becomes theory. One way of saying this is to call a hypothesis falsifiable, that is, when there is evidence which does not fit it. A useful hypothesis, then, is one which has overcome the trials of empirical testing. Theory can then be compared with reality in as many circumstances as possible, before, in some cases, theory becomes law. Although Newton’s Laws of Motion still have that status, the General Theory of Relativity means that at higher velocities, and especially close to the speed of light, there are inconsistencies. Newton’s three laws however still work as well as they always did at the velocities we human beings live at. Motion can be experimentally checked at various velocities to prove this.
Science, however, is not the same as the Scientific Method. Science is a human activity which is based on the use of that method. Science therefore is prone to all the flaws, digressions, accidents and mistakes of human beings, including their own egos. Especially at the cutting edge of Science – fusion, CRISPR, consciousness studies – there are bound to be false starts, disinformation, and people jockeying for recognition and funding. This is not great for Science. One of the most remarkable aspects of the life of James Lovelock was that, declaring himself to be an independent scientist, he was not restricted by corporate law, commercial utility, or by a limited intellectual viewpoint. This allowed him a far more flexible scientific career, in which he made many discoveries essential to humanity. Yet even he was prone to flaws and mistakes. The first iteration of the Gaia Hypothesis (now Gaia Theory, or Earth Systems Science following peer-reviewed experimental proof) was later amended by Lovelock himself owing to several inconsistencies. But that is why Science is so wonderful. It is quite happy to be wrong, because that is for the sake of the better understanding of the world. Science is a human endeavour – a public endeavour.
You would think then that there is little or no hope of verifying or falsifying hypotheses about subjective experiences. This is why some philosophers rely on a combination of Science and Philosophy in order to make their assertions. I can understand why that happens, though I believe it is the wrong path to take. To many, it seems as though use of the Scientific Method is not enough to work with hypotheses about our minds.
Let’s consider again those chimps and human beings. Some people ask: how do I know you are experiencing consciousness? Could you not be a philosophical zombie, apparently conscious because of what is observed of your behaviour, but actually experiencing nothing subjectively? Well, the reason we shouldn’t bother with such thought experiments is that homo sapiens evolved entirely in social circumstances, a situation which had been the case for hundreds of thousands of years into the past even 200,000 years ago, when early homo sapiens appeared in Africa, and which applies to many of our ancestor species, even as far back as homo erectus. We therefore do know that all human beings are like ourselves. We can be certain of that. Because, however, we did not evolve in social groups with chimps, or wolves, or octopi – not to mention AIs – we cannot know for certain that they are conscious as we are. But, having hypothesised all that, a method of enquiry is required before Science can work its wonders.
One possibility is to consider the laws of self-organising systems, which are natural laws. Since animals and conscious human beings come under that aegis, the various laws of self-organisation apply, including the use of energy flow in an open system to overcome the natural direction of entropy – order, in the face of decay into disorder. Three other principles show that the system components (abstract or corporeal) engage in multiple interactions, that there is dynamic nonlinearity which usually includes negative and positive feedback loops, and that there is a balance of the exploration and use of the environment in which the system finds itself. This does not get us too far, however, although it is a central consideration when considering the likelihood of creating artificial consciousness.
As long as the system is modelling the external world based on its sensory information, the opening stage is reached. Such a system will be unconscious, like, say, an amphibian, but it will have a delineated inner world which is self-organising. There will be self and other. This can be verified by observation.
No AI composed of lines of computer code therefore is ever going to become conscious. The writers of innumerable SF novels and films have assumed this is going to happen despite the fundamental flaws in their thinking. Any artificial consciousness – as opposed to an artificial intelligence – must be embodied. It must be a separate, physical self existing in an environment; and it must have needs. Those needs will in due course be felt as subjective feelings.
Some people think that considering a single such artificial consciousness is sufficient to probe how using the Scientific Method might work here. An article in a recent edition of Psychology Today listed the four dominant hypotheses of consciousness: Higher-order Theories, Global Workspace Theories, Integrated Information Theory, and Predictive Processing Theory. None of the summaries of those four offerings mentioned that conscious minds exist only in one circumstance: within groups of other conscious minds. Each summary either implied or directly mentioned “the brain” as the venue for consciousness.
When speaking about this to interested parties in my career as an author, I have used a rough analogy to challenge the viewpoint of the audience. I place my hand in the air between myself and a member of the audience and tell them that is where consciousness is. Sometimes I use another rough analogy, that of dolphins swimming in the ocean. Dolphins can only exist in the sea, which is their medium, in which they survive. Consciousness could be likened to a medium, a social medium, without which human beings cannot survive. We are dolphins in consciousness which exists supported by society. These are rough analogies, but they are valuable in taking away the emphasis from single, solitary brains.
When considering the next level of the evolution of consciousness, therefore, we must always think in terms of groups. A system which, using a mind generated by the self-organisation of its internal systems, grasps that others of its kind are like itself – and that it can use itself as an exemplar to predict and understand the behaviour of others – is well on the way to becoming conscious. So: any conscious animal or artificial system must be embodied, exist in a group, have multiple needs and internal regulation systems, be well endowed with sensory equipment, and be incapable of directly linking to others of its kind. All that can be ascertained by observation.
A system that in the above scenario is seen to show behaviour which indicates it is assessing the felt value of its experiences – positive, like satiation of hunger, or negative, like pain – may have internal feeling, subjectively felt: affect, as it is sometimes called. It may grasp that only it can feel its own such experiences. True conscious experience, after all, is always felt. If it is unconscious – like autonomous body functions, much of visual perception, some types of learning by experience – it is not felt. And if the system can grasp that it can feel such things, it will understand that others in its group should do also.
Such a system would have no defined goal other than to survive and, if a reproductive method exists, to create offspring. That is what animals do in the natural environment: they survive to have young. And although that does count as a goal, it is nowhere near as precise as the various goals created for AIs, all of which are highly specific.
Consciousness is part of human survival – it evolved for a reason. We know that, because we members of homo sapiens survived for around 200,000 years to become what we are today. Consciousness was vital for our survival and success.
Of course, we do still face the problem of knowing whether or not feelings in systems such as those described above are actually being used to further survival. Do such minds experience qualia? The orangey taste of an orange, the greenness of a leaf, the sound of a violin…
Any hypothesis that an animal or artificial system is conscious must be available for interrogation by Science. Even if fabulously complex artificial systems which reproduce in some way can be built, they still need to have evolved in groups – and in unpredictable environments in which they must prioritise all their various needs – for consciousness to become a required attribute. We need, then, to identify their correlates of the representation of consciousness; and that could be achieved by empirical means.
Indeed, for each of our hominid ancestors, as with human beings today, the real unpredictability was the behaviour of everyone else in their social group. The profoundly flexible, variable, often chaotic behaviour of others was the uncertain environment in which our minds evolved. Physical uncertainty – climate change, for instance – was minor in comparison. In social groups, we represented our individual ideas of the world. We made mental bridges using phenomenal experience.
Science, then, when studying consciousness, should be looking for observable evidence of a particular kind. It should be especially looking for the love of experiencing qualia; that is, an impulse to experience all the subjective feelings that we feel and – as with good food, fine wine, or wonderful music – so thoroughly enjoy. We need those feelings and those creative expressions in order to survive and remain sane. That need and its response literally keeps us sane as individuals. Sanity, after all, is the mental equivalent of physical survival. Moreover, sanity is observable, and can be analysed.
Because human beings are born into individual, unique, separated bodies with access to a huge range of sensory experience, they love the feel of life. It is that quality in particular which can be sought out, observed and worked up into a hypothesis by Science. Any conscious system should also possess theory of mind, feel empathy for others of their kind, and be able to negotiate the complexities of social living.
I believe this is something to which Science can apply itself. I think saying that there are aspects of human consciousness which fall beyond the limits of Science is a cop-out. Explanation leads to understanding, and we are creatures who wish to explain everything, not least ourselves. Science gives us a description of reality, not a description of our imaginations. We have to write our own stories, including new stories, and although we are special we should not pretend the human race is so special we each contain some component beyond reach. Any new stories we write about ourselves must be founded on reality.