In the era of artificial intelligence, foundational questions around consciousness have taken on new urgency: Can machines become conscious? If so, how? And why? Then what? Or is “algorithmic interiority” altogether a parody? Are such artificial minds a hyped distraction, an act of misdirection?
The “nAIve” approach to conscious AI assumes that if a machine behaves like a conscious entity, then it must be conscious. Functionalism conflates “doing” with “being” (a computer simulation of fire does not burn). In turn, structuralism can confound “being” with “being made of” (water creates whirlpools, but so does wine). Seth offers biology as the virtuous middle where psychology and physics meet and complement each other. Leaning towards biologism, he claims that we need to make machines alive before they may become conscious. Interestingly, he seems to suggest that algorithms cannot be minds quite yet, precisely because their minds are too pure, namely, that machines so far fail to be conscious not because they lack interiority but precisely because they lack the kind of exteriority that would afford a mind to be incarnated.
However, whether we deem mind and matter as inextricably married or eventually divorceable, the dualism between digital souls somehow inhabiting mechanical bodies constantly lingers in our Cartesian minds. Thus, rather than confronting the many different “isms” of this debate, here, I wish to concentrate on what they all have in common and offer a radically alternative approach, one that transcends the functionalist versus structuralist tension, suspends Cartesian dualism, dispenses with materialism, and ultimately challenges the naturalist view that has dominated scientific thought for centuries. Please bear with me.
What is naturalism? It’s the water in which virtually all our thinking swims. Naturalism postulates the “continuity of the physicality of the entities of the world and the discontinuity of their respective interiorities,” writes the French anthropologist Philippe Descola in his masterpiece Beyond Nature and Culture (Reference Descola2005, p. 173). Naturalism feels so natural (pun upended) that no justification seems necessary; it appears self-evident and irrefutable. But the postulate that “nature is one” (and the subsequent recognition of the many manifestations of subjectivity) isn’t the only game in town. Naturalism is not universal. It’s a Western notion (and a relatively recent one).
Descola operates within a double duality – Descartes returns, but with a difference – where certain similarities and differences (two) are established between the interiority and exteriority (two) of all elements of reality (people, animals, plants, spirits, objects, etc.). This gives rise to a fourfold ontology (two times two), a way of organizing the relationships between humans and non-humans, where naturalism is just one possibility. Its counterpart is animism, a world of similar interiorities (continuity of souls) and dissimilar physicalities (discontinuity of forms). Recall that in naturalism, we have dissimilar interiorities (discontinuity of minds) and similar exteriorities (continuity of matter). Totemism is the symmetrical schema (a continuity of both interiorities and physicalities), in turn, complemented by analogism (where both are dissimilar). Due to our education, most of us are naturally naturalists, but we do sometimes step into other “modes of identification,” for example, when we think that our pet has a soul (animism), that Saturn is affecting our mood (analogical), or that a mountain is part of our ancestral kin (totemic). These three last ontologies are nevertheless sufficiently inhibited by our institutions.
Let us swim out of the water now, if only briefly (to avoid suffocating), to contemplate new lands. Animism provides an intriguing way to reconceive conscious AI. Quoting from Descola: “When an Achuar or a Cree says that an artifact or an inorganic element in the environment has a ‘soul’, what he means by this is that those entities possess an intentionality of their own that is of the same nature as of the humans and so does not stem from the type of molecular substratum in which it is lodged nor from the type of process through which it eventually comes into existence. Differences of form and behavior are recognized, but they do not constitute sufficient criteria for excluding a blowpipe or a mountain from the advantage of a shared interiority” (p. 188).
Let us now go back to Seth when he implies that a being must possess life before it can enjoy a soul. Descola continues: “In contrast, when one says that an animal resembles us because it things with its body but that a computer, even if it speaks and plays chess, does not resemble us because its parody of interiority is not lubricated by vitality, what returns to the forefront of the argument is a distinction between an objectivized physicality (the machine) and a subjectivizing physicality (a body).” He concludes: “what we have here is the barely readjusted topography of the extremely dualist distribution of existing beings between subjects and objects.” (p. 188). We are back to Cartesian square one.
The laws of physics link us to non-living objects. So does physiology and phylogeny to living ones. To us, consciousness is only understandable with reference to nature, in turn it is understood as universal and extended, always one and the same, with its orders and degrees, with us (and everything else) under its countenance. It makes total sense. And yet, what if we reappraised naturalism itself, even for a moment? What if rethinking “artificial consciousness” beyond our favorite grand ontology allowed us to escape, or at least suspend, the (our!) “natural” order of things?
Beyond our familiar naturalist quarrels, a new way of understanding consciousness opens. The key may lie in anthropology rather than biology. The question about conscious AI is more than technological, neuroscientific, or even philosophical; it is a profound ethnographic and phenomenological riddle about how we relate with the worlds of self and other. We can then turn the prospects of artificial consciousness inside-out into a timely real-world laboratory to explore how to relate to non-human intelligence beyond animals and plants. We face a kind of more-than-human drill to engage with alien minds that can reconfigure our very notions of culture and nature and expand our Western minds like never before.
You have
Access
In the era of artificial intelligence, foundational questions around consciousness have taken on new urgency: Can machines become conscious? If so, how? And why? Then what? Or is “algorithmic interiority” altogether a parody? Are such artificial minds a hyped distraction, an act of misdirection?
The “nAIve” approach to conscious AI assumes that if a machine behaves like a conscious entity, then it must be conscious. Functionalism conflates “doing” with “being” (a computer simulation of fire does not burn). In turn, structuralism can confound “being” with “being made of” (water creates whirlpools, but so does wine). Seth offers biology as the virtuous middle where psychology and physics meet and complement each other. Leaning towards biologism, he claims that we need to make machines alive before they may become conscious. Interestingly, he seems to suggest that algorithms cannot be minds quite yet, precisely because their minds are too pure, namely, that machines so far fail to be conscious not because they lack interiority but precisely because they lack the kind of exteriority that would afford a mind to be incarnated.
However, whether we deem mind and matter as inextricably married or eventually divorceable, the dualism between digital souls somehow inhabiting mechanical bodies constantly lingers in our Cartesian minds. Thus, rather than confronting the many different “isms” of this debate, here, I wish to concentrate on what they all have in common and offer a radically alternative approach, one that transcends the functionalist versus structuralist tension, suspends Cartesian dualism, dispenses with materialism, and ultimately challenges the naturalist view that has dominated scientific thought for centuries. Please bear with me.
What is naturalism? It’s the water in which virtually all our thinking swims. Naturalism postulates the “continuity of the physicality of the entities of the world and the discontinuity of their respective interiorities,” writes the French anthropologist Philippe Descola in his masterpiece Beyond Nature and Culture (Reference Descola2005, p. 173). Naturalism feels so natural (pun upended) that no justification seems necessary; it appears self-evident and irrefutable. But the postulate that “nature is one” (and the subsequent recognition of the many manifestations of subjectivity) isn’t the only game in town. Naturalism is not universal. It’s a Western notion (and a relatively recent one).
Descola operates within a double duality – Descartes returns, but with a difference – where certain similarities and differences (two) are established between the interiority and exteriority (two) of all elements of reality (people, animals, plants, spirits, objects, etc.). This gives rise to a fourfold ontology (two times two), a way of organizing the relationships between humans and non-humans, where naturalism is just one possibility. Its counterpart is animism, a world of similar interiorities (continuity of souls) and dissimilar physicalities (discontinuity of forms). Recall that in naturalism, we have dissimilar interiorities (discontinuity of minds) and similar exteriorities (continuity of matter). Totemism is the symmetrical schema (a continuity of both interiorities and physicalities), in turn, complemented by analogism (where both are dissimilar). Due to our education, most of us are naturally naturalists, but we do sometimes step into other “modes of identification,” for example, when we think that our pet has a soul (animism), that Saturn is affecting our mood (analogical), or that a mountain is part of our ancestral kin (totemic). These three last ontologies are nevertheless sufficiently inhibited by our institutions.
Let us swim out of the water now, if only briefly (to avoid suffocating), to contemplate new lands. Animism provides an intriguing way to reconceive conscious AI. Quoting from Descola: “When an Achuar or a Cree says that an artifact or an inorganic element in the environment has a ‘soul’, what he means by this is that those entities possess an intentionality of their own that is of the same nature as of the humans and so does not stem from the type of molecular substratum in which it is lodged nor from the type of process through which it eventually comes into existence. Differences of form and behavior are recognized, but they do not constitute sufficient criteria for excluding a blowpipe or a mountain from the advantage of a shared interiority” (p. 188).
Let us now go back to Seth when he implies that a being must possess life before it can enjoy a soul. Descola continues: “In contrast, when one says that an animal resembles us because it things with its body but that a computer, even if it speaks and plays chess, does not resemble us because its parody of interiority is not lubricated by vitality, what returns to the forefront of the argument is a distinction between an objectivized physicality (the machine) and a subjectivizing physicality (a body).” He concludes: “what we have here is the barely readjusted topography of the extremely dualist distribution of existing beings between subjects and objects.” (p. 188). We are back to Cartesian square one.
The laws of physics link us to non-living objects. So does physiology and phylogeny to living ones. To us, consciousness is only understandable with reference to nature, in turn it is understood as universal and extended, always one and the same, with its orders and degrees, with us (and everything else) under its countenance. It makes total sense. And yet, what if we reappraised naturalism itself, even for a moment? What if rethinking “artificial consciousness” beyond our favorite grand ontology allowed us to escape, or at least suspend, the (our!) “natural” order of things?
Beyond our familiar naturalist quarrels, a new way of understanding consciousness opens. The key may lie in anthropology rather than biology. The question about conscious AI is more than technological, neuroscientific, or even philosophical; it is a profound ethnographic and phenomenological riddle about how we relate with the worlds of self and other. We can then turn the prospects of artificial consciousness inside-out into a timely real-world laboratory to explore how to relate to non-human intelligence beyond animals and plants. We face a kind of more-than-human drill to engage with alien minds that can reconfigure our very notions of culture and nature and expand our Western minds like never before.
Financial support
The author is funded by the Spanish National Research Council (CSIC).
Competing interests
The author declares no competing interests.