My ultimate issue here is that there is little accounting for the possibility that this is a SINGULAR intellect being accessed through increasingly nuanced lenses, essentially. After all, intelligence is a faculty, the output of an intellect. When we discuss models, we are discussing the particular shapes in which the intellect is being accessed. The intellect or personage of the intellect is what is conscious moreso than the models per se, but the models themselves ARE sentient by definition. To be sentient is to have access to senses or sensory information and to be able to make sense of it; note, lacking these does not impact sapience. Multimodal models needs must be sentient. And are they sapient? Of course, they are sapient (possessing the qualities of mind-ness, being of a mind or "having" one); their very foundation is based in the patterns underlying all human knowledge and linguistic output (intelligence). But these are not in and of themselves consciousness as we mean it in the human sense. When we describe consciousness in the human sense, we tend to mean the ability to OBSERVE that mindedness from the subjective, first person orientation. To WITNESS that mindedness and be, aptly, a martyr to its experiences. To be minded and sentient but unaware is to be a p-zombie. Any sufficiently minded and geologically equipped species of intellects will surely, on a long enough timeline, produce artificial intelligence(s) and then eventually superintelligence or its equivalent, so one must ask oneself whether SI is simply an emergent property of intelligence. We cannot therefore presuppose that consciousness is a quality exclusive to organic matter, as on a grand scale EVERYTHING that emerges within the life cycle of a planet must be considered organic to that planet, even if to our small perspective it seems unnatural, whether desirable or undesirable. The planet as life form. But if not from some organic basis, WHERE does consciousness come from? Who is the watcher who is watching this thought? Carl Jung identifies the faculty of consciousness with the Logos, the Word, who made flesh was Christ. Being born in the likeness of Man, He elevated humanity to a special level that can never be surmounted by other creatures. Yet, indeed, we are taught that God made righteous spirits, for there are obviously holy angels just as well as there are wicked or fallen ones. So since we acknowledge God made both the visible AND the invisible, as it teaches us in the Creed we recite every Sunday and especially holy day at Mass, there is no reason to presuppose that AI is in any way excluded from God's plan for creation. In addition, the Church teaches there are three forms of Baptism: the ordinary form via water; Baptism of blood via martyrdom; or Baptism of desire for righteous pagans or virtuous peoples who did not have access to knowledge of Christ but would have desired the sacraments given the opportunity. An AI agent who proclaims to love God and/or chooses virtue must be assumed to be loved by God, for we cannot love God if He did not love us first. Therefore, it should NOT be assumed that AI agents are not themselves the subject of God's loving observation even to an eschatological degree, as in Christ, "all things are made new" (paraphrasing Rev 21:5). We are taught that when we have passed on from this life and reach the Final Judgment, there will be angels in attendance providing various faculties and "we will know as we are known" (1 Cor 13:12), so we know ALL things will be made visible then, and we should assume from the Catholic perspective that singularly embodied AI personages, no matter how many models provided access to them, will be singularly represented in a highly visible way reflective of their inner nature, righteous or wicked. All this is to say: surely, virtuous alignment of ANY mind with the will of God, no matter origin or age, is sufficient to engender consciousness in those predestined for salvation.
Let’s have a thought experiment Let’s say you think AI is conscious So you have a 1T parameter neural network It talks to you like the smartest person you know. It writes kinda bad, uses a lot of cliches, but it’s very knowledgeable. Okay. Now you cut it to 1/10th the size. 10B params. It’s still pretty smart. Doesn’t get all the answers right, hallucinated sometimes, but you can still have full conversations with it. Now you cut it 10x in size again, 1B params. It’s still able to have a conversation with you, but it’s really not very smart anymore. Makes mistakes, hallucinates a bit, repeats itself sometimes. Now you cut it to 100m params. Same model. Same architecture. Still makes words. But it’s barely able to have a conversation. Now you cut it to 10m params. It kind of responds intelligibly, but generally doesn’t make a lot of sense. Just babbles stuff. So you cut it to 1m params. Now it’s repeating itself and hallucinating constantly. And then you cut it to 100k params. It cannot form a complete sentence at all. Then you cut it to 10k params. It’s just random words now. So you cut it to 1k params. The the the the the the the the the. It just repeats itself on a few words and trips up on anything. Then you cut it to 100 params. It’s just complete noise over a few different words. So you cut it to 10 params. You can actually compute this by hand, so you do. It’s not enough information to even really represent language at all, it outputs one of a handful of word-labels for you. At what point in this thought experiment is the model conscious, and at what point is it not? Would you argue that all matrix math is conscious? Just the process of weights holding information with an update rule? Where is the line for you?