Apple's Machine Learning Research Group recently released a paper that outlines the current “inference” capabilities of AI. The details of this study are quite technical from an amateur's perspective, but the results are clear. AI is not as close to actual “intellect” as many people would hope or fear.
Apple's research has investigated cutting-edge large-scale inference model (LRMS), a form of AI that aims to advance further than the large-scale language model (LLMS) that surpasses Openai's widely used ChatGPT. In short, this study found that LRMS exhibits superiority over LLM in specific intermediate complex tasks, but both LRMS and LLMS experience a “complete disintegration” when faced with more complex tasks. As AI content creator and entrepreneur Ruben Hasid summed up the findings, the accuracy of the model ultimately reduced to 0% on the high complexity task, even when researchers gave LRMS an accurate solution algorithm for a particular problem.
AI presents a set of social challenges that are rarely expertise or willing to deal with. The entire industry is afraid of evacuation of jobs, and many are worried about restructuring the global economic order. The government is reluctant to provide regulatory guardrails. The detailed AI 2027 report outlines scenarios that could have caused fear and inexplicable controversy. Higher education is facing a crisis of students rampaging fraud with AI. The newly elected Pope Leo XIV explained the Pope's choice of name in regards to the “industrial revolution” caused by AI (Leo XIII defended the rights of workers during the original Industrial Revolution). Some openly warn of extinction of the species if we don't stop everything.
But perhaps the most pressing challenge posed by AI is what Apple's research highlights. It is the uniqueness and dignity of humans who reason about living things. Both AI enthusiasts and detractors are concerned with issues of consciousness. Can AI achieve the ability that AI self-reflects, reasones and intends? While some AI developers are very certain that AI awareness is possible (and perhaps already recognized), they can only arrive at such conclusions through a set of disastrous false metaphysical assumptions. They assume, first of all, that consciousness is simply an urgent property beyond vast amounts of data and computing power. They further assume that this is how consciousness works even in humans. We are extremely advanced brains that have achieved the ability to think about thought and have the ability to have world first-person perception as an evolutionary accident. Assuming that, since we reduced humans to machines, machines could become human or at least become human in their ability to reason is a small step.
However, there is an infinite qualitative groove between the machinery of calculation and the linguistic predictions on the one hand, and on the other hand, infinitely qualitative groove. Even if we admit that the two are fully integrated (as all classic philosophical positions regarding mental and physical problems are readily recognized), the mind cannot be reduced to the brain. However, the mind precisely transcends neurobiology through first-person experiences of the self and the world. Even if I could thoroughly map my brain's nerve activity during morning coffee, I still can't access my first-person experience of bean flavor notes.
Something is always omitted in the improvised, purely material explanation of the senses, that is, the conscious perception of the phenomenon of experience. Of course, Christians have always been aware of this. We believe in some kind of insignificant soul (theologians have cashed out the exact relationship between the soul and the body in many ways). However, even many nontheistic philosophers realize that mental states cannot be easily reduced to physical states. Consciousness remains a mystery, but it is the mystery that we all experience with the simplicity of our own existence. As theologian David Bentley Hart makes it clear, “electrochemical events may be inseparably related to thought, not thought, but thought, and the empirical inventory of such events does not reveal the content or experiential quality of ideas, desires, wills, or other spiritual events.”
Ultimately, Apple's research reveals what we already know. The machine cannot be thought of. They do not present reason, self-awareness or intentionality. Only humans who are unique among the material matters of the world, and only humans have the idea, intention, hope, and reason. Theologically, this is linked to our personality. We are “who,” “not something,” but made for union and fellowship with God's man. What AI can do is generate very advanced simulations of these things. This does not mean that AI is harmless. It brings true promises and dangers of humanity. However, given the nature of consciousness in philosophical and theological perspectives, we should alleviate the fear that AI can replace the dignity and beauty of human existence.
