Layne Hartsell in discussion with computer scientist Oliver Sampson
We are standing at a peculiar crossroads. Machines now write poetry, compose symphonies, pass university exams, and engage in conversation with unsettling fluency. As artificial intelligence masters tasks we once considered uniquely human, a profound question emerges: what exactly separates us from our creations?
Layne Hartsell says, “but here's the issue: how can we even formulate meaningful questions about mind and human nature when our primary investigative tool—language—is now being masterfully mimicked and we were already confronted with limitations to our understanding, sensory, cognitively, imaginatively”. For language, computer scientist Oliver Sampson calls it "spicy autocomplete"? This is not language. Sampson says “Large language models are sophisticated token prediction devices, statistical engines that reflect the patterns of human culture without possessing any theory of mind themselves. They don't intuit whom they're speaking to or understand the meaning behind their outputs.”
Layne Hartsell says this reveals something crucial: we're the ones projecting theory of mind into these machines and believing it. Like seeing faces in clouds, we attribute intention and understanding to systems that are fundamentally performing probability calculations. As Hartsell and colleague Sampson argue in their pushback against the AI hype machine, we're confusing performance with comprehension. Yet something more insidious may be occurring. Hartsell suggests we're witnessing the further "technification of the human being"—a process where human nature itself is being redefined in computational terms, as is reality. When we measure intelligence as information processing, creativity as novel recombination, and even emotion as predictable patterns, we're not just building machines in our image. We're being forced to match our machines.
Consider Alan Turing's famous test: if a machine can convince us it thinks, does it matter whether it actually does? Turing sidestepped the philosophical quagmire, but perhaps we shouldn't; we should go with Turing’s other response when questioned about intelligent machines “do we say submarines swim? In other philosophical musings, we can see that Kant's triangle exists as a concept independent of any physical manifestation—it's an abstraction our minds grasp. Can an LLM truly grasp abstractions, or does it merely manipulate symbols according to learned patterns producing whatever coordinates are available for such a triangle? And like Russell's paradox exposed contradictions in formal logic, AI may be exposing the limits of our own frameworks for understanding intelligence.
Hartsell notes that embodiment and consciousness remain beyond artificial systems—and probably beyond our natural understanding. AI doesn't navigate the world with a body that hungers, fears, and desires. It doesn't wake at 3 a.m. questioning its existence. These aren't mere details; they may be prerequisites for genuine human experience. Yet as we reduce human cognition to computational processes, we risk losing sight of these irreducible qualities.
Instead of saying that a machine has memory, using human qualities to metaphorically understand machine process, the technification of human nature represents a subtle reductionism: we're adopting the machine's metaphors for ourselves as if we, and Nature, are machines. We speak of "processing" emotions, "bandwidth" for attention, and "optimizing" our lives. This linguistic shift isn't neutral—it shapes how we understand our own humanity, flattening the dynamic, embodied, contradictory experience of being human into something more machine-legible, making human beings susceptible to control.
Yet this critique doesn't render AI powerless. When steered by capable, skilled humans, these systems become formidable tools—amplifiers of human skills, intention, and production rather than replacements for human thought. The danger lies in our willingness to abdicate critical thinking to it, to mistake its confident outputs for wisdom, and most critically, to remake ourselves in its image.
What remains uniquely ours? The question itself. Our capacity to wonder, to doubt our own projections, to resist reduction. To recognize that spicy autocomplete—however impressive—is still just autocomplete, and that we are something far stranger and more profound than any algorithm can capture. In building systems that mirror our linguistic patterns, we're learning that language alone doesn't constitute mind. Consciousness, embodiment, genuine understanding, the irreducible strangeness of subjective experience—these remain human aspects, inseparable, integral.
Oliver Sampson, PhD, CEO and Co-founder of Gitterdan.AI. Sampson works as a Computer Scientist and Adjunct Professor of Machine Learning and AI. He also plays bass and is a Punnerator.
Layne Hartsell, (USA) 雷恩∙哈特塞尔 - 마이클레인핫셀 - 3E: Energy, Economy, Environment - research fellow at the Center for Science, Technology, and Society, Department of Philosophy, Chulalongkorn University in Bangkok, and researcher in Bioregional Letters at North Valley College, Mindanao, Philippines, and also 3E at the Asia Institute, Tokyo. Past affiliations: assistant professor, convergence studies, Sookmyung Women's University and Research Institute for Asian Women - Asia-Pacific Women's Information Network Center in Seoul. Research professor, Sungkyunkwan University and the Advanced Institute of Nanotechnology, Seoul/Suwon. Lecturer at Mahidol University, Siriraj Medical Center, Department of Molecular Biology and Bioinformatics, Bangkok. Researcher at the University of Virginia College of Medicine, Charlottesville. He is a member of the editorial board at Korea IT Times.
Korean version available here:
