Given the complexity of mindedness and our very limited understanding of what constitutes it, what else can we talk about? We could consider further possibilities of artificial systems that might arise, thereby enlarging the set of constraints that has to be satisfied. Some of them seem much more likely than artificial minds, and they might precede minds chronologically. I would like to focus on the idea of fragmented minds on the one hand and of postbiotic systems on the other, as two versions of artificial systems. An artificially-constructed fragmented mind is characterized by only partial satisfaction of the constraints fulfilled by a human mind. It could thus, very much like autistic persons with savant syndrome (i.e., more than average competence in a certain domain, e.g., language learning or music), and possess only some of our cognitive functions, but be strikingly better at them than normal humans are and ever could be, given their biological endowment.[1] Postbiotic minds, on the other hand, could satisfy additional constraints that are not yet apparent presently. I will conclude with some reflections on the new kind of ethics that will have to be created in order to approach new kinds of cognitive agents. As pointed out above, I assume that cognitive agents will be possible much earlier than truly minded agents. Learning, remembering, and other cognitive functions can already be recreated in artificial systems like Spaun. Still, human cognition is very versatile and complex. A fully minded agent, in contrast to a merely cognitive agent, might also be able to experience herself as a cognitive agent.
Therefore, I propose that cognitive systems could be created that do not yet qualify as a copy of our cognitive facilities, but which cover only parts of our cognitive setup. I call these fragmented minds. Importantly, the word minds does not refer here to the artificiality of the system at all. There are human beings with fragmented minds, too, such as babies, who do not yet display the cognitive abilities we ascribe to adult humans in general, or the aforementioned autistic humans with savant syndrome. Fragmented minds are contrasted with what we experience as normal human minds. Fragmented means that the created system possesses only part of the abilities that our mind displays. The term mind delineates the—historically contingent—point of reference that is human beings. How are fragmented minds further characterized? Eliasmith himself gives us an example: we could design a robot (an artificial mind) that gains fulfillment from serving humans (this collection, p. 11). This would only be possible if aspects of our own minds were not part of the mental landscape of this robot. We could roughly formulate such an aspect, such as the will to design one’s own life. Folk psychology would most likely regard this robot as lacking a free will, which is in conformity with the idea of slavery that Eliasmith acknowledges (ibid.). So a fragmented mind is an artificial system that possesses part of a biological cognitive system’s abilities instead of the rich landscape most higher animals (e.g., some fishes and birds, certainly mammals), as well as humans, display.
Related to the aspect of fragmented minds is the idea that we could refrain from creating minds that might cause us a lot of moral and practical trouble, and instead focus on building sophisticated robots designed to carry out specific kinds of tasks. Why do we need to create artificial minds? What is the additional value gained? If these robots are not mindful, we will circumvent the vast majority of conceptual and ethical problems, such as legal questions (What is their legal status compared to ours?) or ethical considerations (If I am not sure whether an artificial agent can perceive pain, how should I treat it in order to not cause harm?). In which case, they would only be more capable technology than what we know at present, and most likely be of no major concern for the philosophy of mind. However, if they are mindful, we doubtlessly have to think about new ways of approaching them ethically.
Also ethically relevant are intermediate systems, systems that are not clearly either natural or artificial. These systems have been called postbiotic systems (Metzinger 2012, p. 268). What characterizes postbiotic systems is the fact that they are made up of both natural and artificial parts, thus belonging to neither of the exhaustive categories “natural” or “artificial”. In that way a natural system, e.g., an animal, could be controlled by artificially-constructed hardware (as in hybrid bio-robotics); or, in the opposite case, artificial hardware could be equipped with biologically-inspired software, which works in very much the same way as neuronal computation (Metzinger 2012, pp. 268–270; Metzinger 2013, p. 4). Perhaps Eliasmith’s own brain-like model Spaun is a postbiotic system in this sense, too. In what way would these systems become ethically relevant? Although the postbiotic systems in existence today do not have the ability to subjectively experience themselves and the world around them, they might have it in the future. In being able to subjectively experience their surroundings, they are probably also able to experience the state of suffering (Metzinger 2013, p. 4). Everything that is able to consciously experience a frustration of preferences as a frustration of its own preferences automatically becomes an object of ethical consideration, according to this principle. For such cases, we have to think of ethical guidelines before we are confronted with a suffering postbiotic mind, which could be much earlier than we expect. Before thinking about how to implement something as complex and unpredictable as an artificial mind, one should consider what one does not want to generate. This could, for example, be the ability to suffer, the inability to judge and act according to ethical premises, or the possibility of developing itself further in a way that is not controllable by and potentially dangerous for humans.