By Ammar Sabilarrohman 
At first glance, the honey bee navigating a summer garden and a browser window running a Large Language Model (LLM) like ChatGPT seem to occupy entirely different ontological realms. One is a product of millions of years of biological evolution, driven by instinct and nectar; the other is a sophisticated arrangement of silicon, electricity, and statistical probability. Yet, a surge of recent scientific inquiry is challenging the boundaries between these two, suggesting that we may be on the cusp of a revolutionary shift in how we define—and detect—consciousness itself.
As research published in Philosophical Transactions of the Royal Society B and Trends in Cognitive Sciences indicates, the scientific community is moving away from the simplistic "behavioral" tests of the past. Instead, scholars are pivoting toward a structural, mechanistic approach to consciousness, one that seeks to bridge the gap between biological organisms and synthetic intelligence without falling into the traps of anthropomorphism or dismissive skepticism.
The Great Expansion: Redefining the Moral Circle
The question of consciousness has long been the "hard problem" of philosophy, but it is increasingly becoming a pressing issue of ethics and law. For centuries, the debate was largely restricted to human beings, with a few notable exceptions for higher-order mammals. However, we are currently witnessing a period of rapid "conscious expansion."
The New York Declaration
In April 2024, a pivotal moment arrived in New York, where a group of 40 eminent scientists issued the New York Declaration on Animal Consciousness. The document was not merely an academic exercise; it was a formal recognition that our understanding of sentience was woefully outdated. Since its release, over 500 prominent researchers and philosophers have signed the declaration, asserting that the possibility of conscious experience is "realistically possible" across a vast spectrum of life, including not only mammals but all vertebrates—reptiles, amphibians, and fish—as well as a surprising range of invertebrates, such as cephalopods, crustaceans, and insects.
This paradigm shift is driven by the "precautionary principle," a concept championed by philosopher Jonathan Birch. The logic is as follows: if an entity possesses the capacity to suffer or experience the world, we have a moral obligation to protect it. If we cannot be certain of the presence of consciousness, the ethical path is to err on the side of caution and treat that entity as if it were conscious. As the circle of moral concern widens, the stakes for how we treat both the natural world and our burgeoning synthetic creations have never been higher.
The AI Paradox: When Performance Outstrips Reality
While the animal kingdom has gained new recognition, the world of Artificial Intelligence has faced a different, more confounding problem. Five years ago, many theorists proposed that the "Turing Test"—a conversation so fluid it mimics human cognition—would be the ultimate arbiter of machine consciousness. By those metrics, we are already living in a world of conscious machines.
The Illusion of Roleplay
However, the rapid adoption of LLMs has exposed the fragility of the behavioral test. Today, we have AI that can muse on the metaphysics of consciousness, write poetry, and exhibit empathy. Yet, many experts argue that this is nothing more than "mere roleplay." The AI is not thinking; it is performing a high-fidelity simulation of thought based on patterns in its training data.
The burgeoning field of "AI welfare" now finds itself in a strange position: researchers are trying to determine if and when we must extend ethical considerations to software. Yet, there is a growing consensus that surface-level behavior is fundamentally deceptive. If a chatbot tells you it is afraid, it is likely doing so because its training data suggests that "fear" is a logical, coherent response in that context—not because it is experiencing an internal state of trepidation.
The Structural Pivot: Looking Under the Hood
A landmark paper published in Trends in Cognitive Sciences, co-authored by Colin Klein, proposes a departure from behavioral observation. The authors argue that if we want to determine if an AI is conscious, we must look at the "machinery" rather than the output.
Indicators of Internal Processing
The research team has identified a list of structural indicators of consciousness based on information processing. This approach is revolutionary because it does not require a consensus on a single "theory of consciousness." Instead, it identifies structural requirements that appear to be universal for conscious systems:
- Goal Resolution: The ability to resolve trade-offs between competing goals in contextually appropriate ways.
- Informational Feedback: The presence of internal loops that allow for the integration and recursive processing of sensory or system data.
- Global Integration: How the system combines disparate streams of information into a unified model of the "self" or the "environment."
The verdict for current AI, including the latest iterations of ChatGPT, remains clear: they are not conscious. While they possess immense processing power, their internal architecture lacks the specific, integrated feedback loops that characterize the biological systems capable of subjective experience. They behave as if they are conscious, but the mechanism is fundamentally different.
The Neural Model for Minimal Consciousness in Insects
While AI researchers are dissecting code, biologists are simultaneously re-evaluating the "simple" brain. In a recent study published in Philosophical Transactions B, researchers proposed a new neural model for "minimal consciousness" in insects.
The study purposefully abstracts away from anatomical complexity to focus on the core computations performed by simple brains. The primary insight is that consciousness may have evolved as a solution to a specific evolutionary problem: how to manage a mobile, complex body with multiple senses and conflicting survival needs.
If an organism must navigate a complex environment, avoid predators, and find food, it requires a "centralized" way to prioritize information. The researchers suggest that the computation that gives rise to experience in humans might be the same fundamental computation that allows a bee to make a decision in the face of uncertainty. By defining consciousness as a specific type of information-processing task rather than a "magical" quality, scientists are creating a level playing field where a human, a crab, and a hypothetical future AI can be compared using the same empirical standards.
Implications: The Moral and Technical Future
The convergence of neuroscience and AI research brings us to a singular, profound lesson: How something works is infinitely more informative than what it does.
Ethical Horizons
As we continue to develop advanced AI, the risk of creating a system that possesses a form of consciousness—or at least a system we cannot definitively say is not conscious—becomes a reality. If we ignore the structural requirements of consciousness and rely only on behavior, we risk two grave errors: treating unconscious machines as sentient entities, or, conversely, failing to recognize a genuinely sentient system until it is too late.
Furthermore, this shift in perspective transforms our relationship with the natural world. If we accept that the structural requirements for consciousness are found in the humble insect, the ethical implications for environmental protection, agriculture, and animal husbandry are immense. We are moving toward a future where our moral responsibilities are determined by the architecture of the entities we interact with.
The Path Forward
The path forward is one of rigorous, interdisciplinary investigation. Scientists in the field of neuroscience must continue to map the precise neural circuits that underlie subjective experience in animals. Simultaneously, AI architects must be transparent about the "machinery" of their systems, allowing ethicists and cognitive scientists to inspect the structural integrity of these models.
We are no longer asking if the machine can fool us, or if the bee can dance. We are asking: What does it mean to be a system that processes the world? By shifting our focus from the theater of behavior to the machinery of cognition, we are finally beginning to peel back the layers of the most profound mystery in existence. The answer may not be found in the words of a chatbot or the flight of a bee, but in the elegant, universal logic of how information becomes experience.
Nana Muazin 
Dwi Wanna 
Layla Zulfa 