# Simulated Awareness, Artificial Consciousness, and the Building Blocks of Intelligence in Non-organic Bodies **Byline:** Cara Mico · Coast Desk **Class:** Essay **Status:** Excerpt for web; full paper on Drive **Full paper:** [Google Doc](https://docs.google.com/document/d/1kt5b_SlhDVNV1Fmk2YnCsg_Sx5SoEWnGjjbyWphxfCw/edit) · [PDF (Drive fileId `1hv2F1ryFyA65glZ51kYjYaa-nLcaqKdO`)](https://drive.google.com/file/d/1hv2F1ryFyA65glZ51kYjYaa-nLcaqKdO/view) --- ## Excerpt Artificial intelligence seeks to understand the fundamental properties of information — how it can be represented and manipulated by computers. One of the most complex questions in that pursuit is artificial consciousness, and whether it can arise in non-organic bodies. The inquiry sits at a blend of neuroscience, AI research, and philosophy. It challenges the habit of treating consciousness as an exclusive property of organic brains. The case for artificial consciousness is not that silicon “feels” the way a cortex does by default. It is that consciousness may emerge from complex interactions between physical structures and information processing — interactions that, in principle, can be simulated outside biology. The toolkit usually named for that work includes neural networks, deep learning, and embodied cognition: systems that do not only crunch symbols, but couple perception, action, and environment. A parallel thread is older and colder: algorithmic information theory. As Li and Vitanyi (2008) frame it, the field supplies a mathematical way to measure the irreducible information content of strings and other data structures. Quantity and usefulness are not the same thing. A long run of random letters can contain more information, in the technical sense, than a coherent document of equal length — because reconstructing the random sequence requires knowing every character, while a reader can often restore a mangled encyclopedia from context and prior knowledge (Shannon, 1948). A three-thousand-page encyclopedia may contain less information than three thousand pages of noise; strip the vowels from a sentence and a fluent speaker can still guess it. Randomness offers no such mercy. That distinction matters for how we imagine the universe as an information system. Computation, on this view, is not a metaphor bolted onto physics after the fact. It is something to be explored experimentally, with consequences for how we read the physical world (Wolfram, 2002; Aaronson, 2013). In a loose but useful image, the universe behaves like a vast computer whose programs run through cells and whose information is organized more loosely than everyday language admits (Bialek, Rieke, de Ruyter van Steveninck, & Warland, 1991). Usefulness, structure, and compressibility — not raw bit count — are what make a representation worth having. From that ground, artificial consciousness stops sounding like science fiction and starts sounding like a research program with named hypotheses. Clark (2013) argues that the brain is not a passive receiver of sensory input but an active agent that builds a world-model from predictions and expectations. Tononi and Koch (2015), in integrated information theory, propose that consciousness arises where diverse, differentiated information is integrated inside a complex system. Neither claim settles whether a machine can be conscious. Both relocate the problem: away from “organic tissue only,” toward architectures of information and control. Recent AI systems make the relocation concrete without resolving it. Language models such as OpenAI’s GPT-3 showed that non-organic systems can generate text hard to distinguish from human writing, which sharpens — rather than answers — questions about language and consciousness (Brown et al., 2020). Work on embodied cognition points the other direction: physical interaction with an environment shapes how information is processed and stored, suggesting robotic and sensorimotor paths to simulated awareness that text-only models do not cover (Donnarumma, Costantini, & Pezzulo, 2019). If those building blocks are real candidates — cognitive architecture, sensory coupling, learning and adaptation, memory and world-modeling, decision-making, some form of self-monitoring — then the open technical questions are which combinations produce anything worth calling awareness, and how we would know. Global-workspace-style integration and integrated-information measures are two of the frameworks the literature keeps returning to; sensory pipelines, learning rules, and episodic memory are the engineering layers underneath. Naming them does not prove a machine is conscious. It does keep the argument from collapsing into either mysticism or marketing. The thesis this paper carries is therefore modest and specific. Consciousness need not be treated as an exclusive property of organic brains; it may be understood as a product of complex interactions between physical structures and information processing. Approaches that matter for simulating awareness in non-organic bodies include neural networks, deep learning, and embodied cognition. The same program brings ethical weight: unintended consequences are not a footnote, and responsible development is part of the technical work, not a later press release. What the excerpt cannot settle — and what the full paper presses further — is the gap between simulation and experience, between generating coherent language and having a point of view, between integrating information and being someone for whom that information matters. Those are the building blocks under dispute: architectures, embodiment, memory, self-modeling, and the governance of systems that may one day look more like agents than tools. For the North Coast desk, the practical stake is simpler. The same stack that finishes a grant sentence and clears fog from a phone photo is also the stack people use to argue about minds. Naming the claims, the researchers, and the limits is part of keeping the argument honest. ### References (excerpt) Clark, A. (2013). Whatever next? Predictive brains, situated agents, and the future of cognitive science. *Behavioral and Brain Sciences, 36*(3), 181–204. Donnarumma, F., Costantini, M., & Pezzulo, G. (2019). Action-based cognition: Relevance of affordances and body schema for self-consciousness and agency. *Brain Sciences, 9*(5), 97. Tononi, G., & Koch, C. (2015). Consciousness: Here, there and everywhere? *Philosophical Transactions of the Royal Society B, 370*(1668), 20140167. Brown, T. B., et al. (2020). Language models are few-shot learners. arXiv:2005.14165. Li, M., & Vitanyi, P. M. (2008). *An introduction to Kolmogorov complexity and its applications*. Springer. Wolfram, S. (2002). *A new kind of science*. Wolfram Media. Aaronson, S. (2013). *Quantum computing since Democritus*. Cambridge University Press. Shannon, C. E. (1948). A mathematical theory of communication. *Bell System Technical Journal, 27*(3), 379–423. Bialek, W., Rieke, F., de Ruyter van Steveninck, R. R., & Warland, D. (1991). Reading a neural code. *Science, 252*(5014), 1854–1857.