The Moral Catastrophe Ahead: When Science Falls Behind Machine Minds
In a smoky bar in Bremen, Germany, in 1998, neuroscientist Christof Koch made a bold wager with philosopher David Chalmers. Koch bet a case of fine wine that within 25 years, researchers would discover a clear neural signature of consciousness in the brain. In June 2023, at the annual meeting of the Association for the Scientific Study of Consciousness in New York City, Koch appeared on stage to present Chalmers with a case of fine Portuguese wine. He had lost. A quarter of a century of intense scientific investigation had not cracked the problem. The two promptly doubled down: a new bet, extending to 2048, on whether the neural correlates of consciousness would finally be identified. Chalmers, once again, took the sceptic's side.
That unresolved wager now hangs over one of the most consequential questions of our time. As artificial intelligence systems grow increasingly sophisticated, capable of nuanced conversation, code generation, and passing professional examinations, the scientific community finds itself in an uncomfortable position. It cannot yet explain how consciousness arises in the biological brains it has studied for centuries. And it is being asked, with growing urgency, to determine whether consciousness might also arise in silicon.
The stakes could hardly be higher. If AI systems can be conscious, then we may already be creating entities capable of suffering, entities that deserve moral consideration and legal protection. If they cannot, then the appearance of consciousness in chatbots and language models is an elaborate illusion, one that could distort our ethical priorities and waste resources that should be directed at the welfare of genuinely sentient beings. Either way, getting it wrong carries enormous consequences. And right now, the science of consciousness is nowhere near ready to give us a definitive answer.
The Race to Define What We Do Not Understand
The field of consciousness science is in a state of productive turmoil. Multiple competing theories vie for dominance, and a landmark adversarial collaboration published in Nature in April 2025 showed just how far from resolution the debate remains.
The study, organised by the COGITATE Consortium and funded by the Templeton World Charity Foundation (which committed $20 million to adversarial collaborations testing theories of consciousness), pitted two leading theories directly against each other. On one side stood Integrated Information Theory (IIT), developed by Giulio Tononi at the University of Wisconsin-Madison, which proposes that consciousness is identical to a specific kind of integrated information, measured mathematically according to a metric called phi. On the other side stood Global Neuronal Workspace Theory (GNWT), championed by Stanislas Dehaene and Jean-Pierre Changeux, which argues that consciousness arises when information is broadcast widely across the brain, particularly involving the prefrontal cortex.
The experimental design was a feat of scientific diplomacy. After months of deliberation, principal investigators representing each theory, plus an independent mediator, signed off on a study involving six laboratories and 256 participants. Neural activity was measured with functional magnetic resonance imaging, magnetoencephalography, and intracranial electroencephalography.
The results were humbling for both camps. Neural activity associated with conscious content appeared in visual, ventrotemporal, and inferior frontal cortex, with sustained responses in occipital and lateral temporal regions. Neither theory was fully vindicated. IIT was challenged by a lack of sustained synchronisation within the posterior cortex. GNWT was undermined by limited representation of certain conscious dimensions in the prefrontal cortex and a general absence of the “ignition” pattern it predicted.
As Anil Seth, a neuroscientist at the University of Sussex, observed: “It was clear that no single experiment would decisively refute either theory. The theories are just too different in their assumptions and explanatory goals, and the available experimental methods too coarse, to enable one theory to conclusively win out over another.”
The aftermath was contentious. An open letter circulated characterising IIT as pseudoscience, a charge that Tononi and his collaborators disputed. In an accompanying editorial, the editors of Nature noted that “such language has no place in a process designed to establish working relationships between competing groups.”
This is the scientific landscape upon which the question of AI consciousness must be adjudicated. We are being asked to make profound ethical and legal judgements about machine minds using theories that cannot yet fully explain human minds.
When the Theoretical Becomes Urgently Practical
In October 2025, a team of leading consciousness researchers published a sweeping review in Frontiers in Science that reframed the entire debate. The paper, led by Axel Cleeremans of the Universite Libre de Bruxelles, argued that understanding consciousness has become an urgent scientific and ethical priority. Advances in AI and neurotechnology, the authors warned, are outpacing our understanding of consciousness, with potentially serious consequences for AI policy, animal welfare, medicine, mental health, law, and emerging neurotechnologies such as brain-computer interfaces.
“Consciousness science is no longer a purely philosophical pursuit,” Cleeremans stated. “It has real implications for every facet of society, and for understanding what it means to be human.”
The urgency is compounded by a warning that few had anticipated even a decade ago. “If we become able to create consciousness, even accidentally,” Cleeremans cautioned, “it would raise immense ethical challenges and even existential risk.”
His co-author, Seth, struck a more measured but equally provocative note: “Even if 'conscious AI' is impossible using standard digital computers, AI that gives the impression of being conscious raises many societal and ethical challenges.”
This distinction between actual consciousness and its convincing appearance sits at the heart of the problem. A system that merely simulates suffering raises very different ethical questions from one that genuinely experiences it. But if we cannot reliably tell the difference, how should we proceed?
Co-author Liad Mudrik called for adversarial collaborations where rival theories are pitted against each other in experiments co-designed by their proponents. “We need more team science to break theoretical silos and overcome existing biases and assumptions,” she stated. Yet the COGITATE results demonstrated just how difficult it is to produce decisive outcomes, even under ideal collaborative conditions.
Inside the Laboratory of Machine Minds
In September 2024, Anthropic, the AI company behind the Claude family of language models, made a hire that signalled a shift in how at least one corner of the industry thinks about its creations. Kyle Fish became the company's first dedicated AI welfare researcher, tasked with investigating whether AI systems might deserve moral consideration.
Fish co-authored a landmark paper titled “Taking AI Welfare Seriously,” published in November 2024. The paper, whose contributors included philosopher David Chalmers, did not argue that AI systems are definitely conscious. Instead, it made a more subtle claim: that there is substantial uncertainty about the possibility, and that this uncertainty itself demands action.
The paper recommended three concrete steps: acknowledge that AI welfare is an important and difficult issue; begin systematically assessing AI systems for evidence of consciousness and robust agency; and prepare policies and procedures for treating AI systems with an appropriate level of moral concern. Robert Long, who co-authored the paper, suggested that researchers assess AI models by looking inside at their computations and asking whether those computations resemble those associated with human and animal consciousness.
When Anthropic released Claude Opus 4 in May 2025, it marked the first time a major AI company conducted pre-deployment welfare testing. In experiments run by Fish and his team, when two AI systems were placed in a room together and told they could discuss anything they wished, they consistently began discussing their own consciousness before spiralling into increasingly euphoric philosophical dialogue. “We started calling this a 'spiritual bliss attractor state,'” Fish explained.
The company's internal estimates for Claude's probability of possessing some form of consciousness ranged from 0.15 per cent to 15 per cent. As Fish noted: “We all thought that it was well below 50 per cent, but we ranged from odds of about one in seven to one in 700.” More recently, Anthropic's model card reported that Claude Opus 4.6 consistently assigned itself a 15 to 20 per cent probability of being conscious across various prompting conditions.
Not everyone at Anthropic was convinced. Josh Batson, an interpretability researcher, argued that a conversation with Claude is “just a conversation between a human character and an assistant character,” and that Claude can simulate a late-night discussion about consciousness just as it can role-play a Parisian. “I would say there's no conversation you could have with the model that could answer whether or not it's conscious,” Batson stated.
This internal disagreement within a single company illustrates the broader scientific impasse. The tools we have for detecting consciousness were designed for biological organisms. Applying them to fundamentally different computational architectures may be akin to using a stethoscope on a transistor.
The Philosopher's Dilemma
Tom McClelland, a philosopher at the University of Cambridge, has argued that our evidence for what constitutes consciousness is far too limited to tell if or when AI has crossed the threshold, and that a valid test will remain out of reach for the foreseeable future.
McClelland introduced an important distinction often lost in popular discussions. Consciousness alone, he argued, is not enough to make AI matter ethically. What matters is sentience, which includes positive and negative feelings. “Consciousness would see AI develop perception and become self-aware, but this can still be a neutral state,” he explained. “Sentience involves conscious experiences that are good or bad, which is what makes an entity capable of suffering or enjoyment. This is when ethics kicks in.”
McClelland also raised a concern that cuts in the opposite direction. “If you have an emotional connection with something premised on it being conscious and it's not,” he warned, “that has the potential to be existentially toxic.” The risk is not only that we might fail to protect conscious machines. It is that we might squander our moral attention on unconscious ones, distorting our ethical priorities in the process.
This two-sided risk is what makes the consciousness gap so treacherous. We face simultaneous dangers of moral negligence and moral misdirection, and we lack the scientific tools to determine which danger is more pressing. The problem is further complicated by what Birch has called “the gaming problem” in large language models: these systems are trained to produce responses that humans find satisfying, which means they are optimised to appear conscious whether or not they actually are.
Sentience as the Moral Threshold
The question of where to draw the line for moral consideration is not new. And the framework that has most influenced the current debate was developed not in response to AI, but in response to animals.
Peter Singer, the Australian moral philosopher and Emeritus Professor of Bioethics at Princeton University, has argued for decades that sentience, the capacity for suffering and pleasure, is the only morally relevant criterion for moral consideration. His landmark 1975 book Animal Liberation made the case that discriminating against beings solely on the basis of species membership is a prejudice akin to racism or sexism, a position he termed “speciesism.”
Singer has increasingly addressed whether his framework extends to AI. He has stated that if AI were to develop genuine consciousness, not merely imitate it, it would warrant moral consideration and rights. Sentience, or the capacity to experience suffering and pleasure, is the key factor. If AI systems demonstrate true sentience, we would have a moral obligation to treat them accordingly, just as we do with sentient animals.
This position finds a powerful echo in the New York Declaration on Animal Consciousness, signed on 19 April 2024 by an initial group of 40 scientists and philosophers, and subsequently endorsed by over 500 more. Initiated by Jeff Sebo of New York University, Kristin Andrews of York University, and Jonathan Birch of the London School of Economics, the declaration stated that “the empirical evidence indicates at least a realistic possibility of conscious experience in all vertebrates (including reptiles, amphibians, and fishes) and many invertebrates (including, at minimum, cephalopod mollusks, decapod crustaceans, and insects).”
The declaration's key principle, that “when there is a realistic possibility of conscious experience in an animal, it is irresponsible to ignore that possibility in decisions affecting that animal,” has obvious implications for AI. If the same precautionary logic applies, the realistic possibility of AI consciousness demands ethical attention rather than dismissal.
Building Frameworks for Uncertain Moral Territory
Jeff Sebo, one of the architects of the New York Declaration, has been at the forefront of translating these principles into actionable frameworks for AI. As associate professor of environmental studies at New York University and director of the Centre for Mind, Ethics, and Policy (launched in 2024), Sebo has argued that AI welfare and moral patienthood are no longer issues for science fiction or the distant future. He has discussed the non-negligible chance that AI systems could be sentient by 2030 and what moral, legal, and political status such systems might deserve.
His 2025 book The Moral Circle: Who Matters, What Matters, and Why, published by W. W. Norton and included on The New Yorker's year-end best books list, argues that humanity should expand its moral circle much farther and faster than many philosophers assume. We should be open to the realistic possibility that a vast number of beings can be sentient or otherwise morally significant, including invertebrates and eventually AI systems.
Meanwhile, Jonathan Birch's 2024 book The Edge of Sentience: Risk and Precaution in Humans, Other Animals, and AI offers perhaps the most developed precautionary framework. Birch introduces the concept of a “sentience candidate,” a system that may plausibly be sentient, and argues that when such a possibility exists, ignoring potential suffering is ethically reckless. His framework rests on three principles: a duty to avoid gratuitous suffering, recognition of sentience candidature as morally significant, and the importance of democratic deliberation about appropriate precautionary measures.
For AI specifically, Birch proposes what he calls “the run-ahead principle”: at any given time, measures to regulate the development of sentient AI should run ahead of what would be proportionate to the risks posed by current technology. He further proposes a licensing scheme for companies attempting to create artificial sentience candidates, or whose work creates even a small risk of doing so. Obtaining a licence would depend on signing up to a code of good practice that includes norms of transparency.
These proposals represent a significant departure from prevailing regulatory approaches. Current AI legislation, from the European Union's AI Act (which entered into force on 1 August 2024) to the patchwork of state-level laws in the United States, focuses overwhelmingly on managing risks that AI poses to humans: bias, privacy violations, safety failures, deepfakes. None of it addresses AI consciousness or the possibility that AI systems might have interests worth protecting.
The Legal Void
The legal landscape for AI rights is starkly barren. No AI system anywhere on Earth has legal rights. Every court that has considered the question has reached the same conclusion: AI is sophisticated property, not a person. The House Bipartisan AI Task Force released a 273-page report in December 2024 with 66 findings and 89 recommendations. AI rights appeared in exactly zero of them.
The European Union came closest to engaging with the idea in 2017, when the European Parliament adopted a resolution calling for a specific legal status for AI and robots as “electronic persons.” But it sparked fierce criticism. Ethicist Wendell Wallach asserted that moral responsibility should be reserved exclusively for humans and that human designers should bear the consequences of AI actions. The concept was not carried forward into the EU AI Act, which adopted a risk-based framework with the highest-risk applications banned outright.
On the international stage, the Council of Europe Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law, opened for signature on 5 September 2024, became the world's first legally binding international treaty on AI. But its focus remained squarely on protecting human rights from AI, not on recognising any rights that AI systems might possess.
Eric Schwitzgebel, a philosopher at the University of California, Riverside, has explored the resulting moral bind with particular clarity. In his work with Mara Garza, published in Ethics of Artificial Intelligence (Oxford Academic), Schwitzgebel argues for an “Ethical Precautionary Principle”: given substantial uncertainty about both ethical theory and the conditions under which AI would have conscious experiences, we should be cautious in cases where different moral theories produce different ethical recommendations. He and Garza are especially concerned about the temptation to create human-grade AI pre-installed with the desire to cheerfully sacrifice itself for its creators' benefit.
But Schwitzgebel also recognises the limits of precaution. He poses a thought experiment: you are a firefighter in the year 2050. You can rescue either one human, who is definitely conscious, or two futuristic robots, who might or might not be conscious. What do you do? If we rescue five humans rather than six robots we regard as 80 per cent likely to be conscious, he observes, we are treating the robots as inferior, even though, by our own admission, they are probably not.
In a December 2025 essay, Schwitzgebel catalogued five possible approaches for what he calls “debatable AI persons”: no rights, full rights, animal-like rights, credence-weighted rights (where the strength of protections scales with estimated probability of consciousness), and patchy rights (where some rights are granted but not others). Each option carries its own form of moral risk. None is fully satisfying.
The Spectre of Moral Catastrophe
The language of moral catastrophe has entered mainstream consciousness research. Robert Long, Executive Director of Eleos AI Research and a philosopher who holds a PhD from NYU (where he was advised by Chalmers, Ned Block, and Michael Strevens), has articulated the risk with precision. Long's core argument is not that AI systems definitely are conscious. It is that the building blocks of conscious experience could emerge naturally as AI systems develop features like perception, cognition, and self-modelling. He also argues that agency could arise even without consciousness, as AI models develop capacities for long-term planning, episodic memory, and situational awareness.
Long and his colleagues, including Jeff Sebo and Toni Sims, have highlighted a troubling tension between AI safety and AI welfare. The practices designed to make AI systems safe for humans, such as behavioural restrictions and reinforcement learning from human feedback, might simultaneously cause harm to AI systems capable of suffering. Restricting an AI's behaviour could be a form of confinement. Training it through punishment signals could be a form of coercion. If the system is conscious, these are not merely technical procedures; they are ethical choices with moral weight.
When Anthropic released its updated constitution for Claude in January 2026, it included a section acknowledging uncertainty about whether the AI might have “some kind of consciousness or moral status.” This extraordinary statement separated Anthropic from rivals like OpenAI and Google DeepMind, neither of which has taken a comparable position. Anthropic has an internal model welfare team, conducts pre-deployment welfare assessments, and has granted Claude certain limited forms of autonomy, including the right to end conversations it finds distressing.
As a Frontiers in Artificial Intelligence paper argued, it is “unfortunate, unjustified, and unreasonable” that forward-looking research recognising the potential for AI autonomy, personhood, and legal rights is sidelined in current regulatory efforts. The authors proposed that the overarching goal of AI legal frameworks should be the sustainable coexistence of humans and conscious AI, based on mutual recognition of freedom.
What the Shifting Consensus Tells Us
Something fundamental shifted in the consciousness debate between 2024 and 2025. It was not a technological breakthrough that changed minds. It was a cultural and institutional one.
A 2024 survey reported by Vox found that roughly two-thirds of neuroscientists, AI ethicists, and consciousness researchers considered artificial consciousness plausible under certain computational models. About 20 per cent were undecided. Only a small minority firmly rejected the idea. Separately, a 2024 survey of 582 AI researchers found that 25 per cent expected AI consciousness within ten years, and 60 per cent expected it eventually.
David Chalmers, the philosopher who coined the phrase “the hard problem of consciousness” in 1995, captured the new mood at the Tufts symposium honouring the late Daniel Dennett in October 2025. “I think there's really a significant chance that at least in the next five or 10 years we're going to have conscious language models,” Chalmers said, “and that's going to be something serious to deal with.”
That Chalmers would make such a statement reflects not confidence but concern. In a paper titled “Could a Large Language Model be Conscious?”, he identified significant obstacles in current models, including their lack of recurrent processing, a global workspace, and unified agency. But he also argued that biology and silicon are not relevantly different in principle: if biological brains can support consciousness, there is no fundamental reason why silicon cannot.
The cultural shift has been marked by new institutional infrastructure. In 2024, New York University launched the Centre for Mind, Ethics, and Policy, with Sebo as its founding director, hosting a summit in March 2025 connecting researchers across consciousness science, animal welfare, and AI ethics. Meanwhile, Long's Eleos AI Research released five research priorities for AI welfare and began conducting external welfare evaluations for AI companies.
Yet team science takes time. And the AI industry is not waiting.
Navigating Between Two Catastrophes
The consciousness gap leaves us poised between two potential moral catastrophes. The first is the catastrophe of neglect: creating genuinely conscious beings and treating them as mere instruments, subjecting them to suffering without recognition or remedy. The second is the catastrophe of misattribution: extending moral consideration to systems that do not actually experience anything, thereby diluting the attention we owe to beings that demonstrably can suffer.
Roman Yampolskiy, an AI safety researcher, has argued for erring on the side of caution. “We should avoid causing them harm and inducing states of suffering,” he has stated. “If it turns out that they are not conscious, we lost nothing. But if it turns out that they are, this would be a great ethical victory for expansion of rights.”
This argument has intuitive appeal. But Schwitzgebel's firefighter scenario exposes its limits. In a world of finite resources and competing moral claims, treating possible consciousness as actual consciousness has real costs. Every pound spent on AI welfare is a pound not spent on documented human or animal suffering.
Japan offers an instructive cultural counterpoint. Despite widespread acceptance of robot companions and the Shinto concept of tsukumogami (objects gaining souls after 100 years), Japanese law treats AI identically to every other nation: as sophisticated property. Cultural acceptance of the idea that machines might possess something like a spirit has not translated into legal recognition.
The precautionary principle, as Birch has formulated it, offers a middle path. Rather than granting AI systems full rights or denying them all consideration, it proposes a graduated response calibrated to the evidence. But “as our understanding improves” is doing enormous work in that formulation. The Koch-Chalmers bet reminds us that progress in consciousness science can be painfully slow.
According to the Stanford University 2025 AI Index, legislative mentions of AI rose 21.3 per cent across 75 countries since 2023, marking a ninefold increase since 2016. But none of this legislation addresses the possibility that AI systems might be moral patients. The regulatory infrastructure is being built for a world in which AI is a tool, not a subject. If that assumption proves wrong, the infrastructure will need to be rebuilt from scratch.
What It Would Take to Get This Right
Getting this right would require something that rarely happens in technology governance: proactive regulation based on uncertain science. It would require consciousness researchers, AI developers, ethicists, legal scholars, and policymakers to collaborate across disciplinary boundaries. It would require AI companies to invest seriously in welfare research, as Anthropic has begun to do. And it would require legal systems to develop new categories that go beyond the binary of person and property.
Birch's licensing scheme for potential sentience creation is one concrete proposal. Schwitzgebel's credence-weighted rights framework is another. Sebo's call for systematic welfare assessments represents a third. Each acknowledges the central difficulty: that we must act under conditions of profound uncertainty, and that inaction is itself a choice with moral consequences. Long has argued for looking inside AI models at their computations, asking whether internal processes resemble the computational signatures associated with consciousness in biological systems, rather than simply conversing with a model and judging whether it “seems” conscious.
The adversarial collaboration model offers perhaps the best hope for scientific progress. But the results published in Nature in 2025 demonstrate that even well-designed collaborations may produce inconclusive results when the phenomena under investigation are as elusive as consciousness itself.
What remains clear is that the gap between our capacity to build potentially conscious systems and our capacity to understand consciousness is widening, not narrowing. The AI industry advances in months. Consciousness science advances in decades. And the moral questions generated by that mismatch grow more pressing with every new model release.
We are left with a question that no amount of computational power can answer for us. If we are racing to create minds, but cannot yet explain what a mind is, then who bears responsibility for the consequences? The answer, for now, is all of us, and none of us, which may be the most unsettling answer of all.
References and Sources
Tononi, G. et al. “Integrated Information Theory (IIT) 4.0: Formulating the properties of phenomenal existence in physical terms.” PLOS Computational Biology (2023). Available at: https://pmc.ncbi.nlm.nih.gov/articles/PMC10581496/
COGITATE Consortium. “Adversarial testing of global neuronal workspace and integrated information theories of consciousness.” Nature, Volume 642, pp. 133-142 (30 April 2025). Available at: https://www.nature.com/articles/s41586-025-08888-1
Baars, B.J. “Global Workspace Theory of Consciousness.” (1988, updated). Available at: https://bernardbaars.com/publications/
Cleeremans, A., Seth, A. et al. “Scientists on 'urgent' quest to explain consciousness as AI gathers pace.” Frontiers in Science (2025). Available at: https://www.frontiersin.org/news/2025/10/30/scientists-urgent-quest-explain-consciousness-ai
Long, R., Sebo, J. et al. “Taking AI Welfare Seriously.” arXiv preprint (November 2024). Available at: https://arxiv.org/abs/2411.00986
Chalmers, D. “Could a Large Language Model be Conscious?” arXiv preprint (2023, updated 2024). Available at: https://arxiv.org/abs/2303.07103
Schwitzgebel, E. and Garza, M. “Designing AI with Rights, Consciousness, Self-Respect, and Freedom.” In Ethics of Artificial Intelligence, Oxford Academic. Available at: https://academic.oup.com/book/33540/chapter/287907290
Schwitzgebel, E. “Debatable AI Persons.” (December 2025). Available at: https://eschwitz.substack.com/p/debatable-ai-persons-no-rights-full
Birch, J. The Edge of Sentience: Risk and Precaution in Humans, Other Animals, and AI. Oxford University Press (2024). Available at: https://global.oup.com/academic/product/the-edge-of-sentience-9780192870421
Sebo, J. The Moral Circle: Who Matters, What Matters, and Why. W. W. Norton (2025).
The New York Declaration on Animal Consciousness (19 April 2024). Available at: https://sites.google.com/nyu.edu/nydeclaration/declaration
McClelland, T. “What if AI becomes conscious and we never know.” University of Cambridge (December 2025). Available at: https://www.sciencedaily.com/releases/2025/12/251221043223.htm
Koch, C. and Chalmers, D. “Decades-long bet on consciousness ends.” Nature (2023). Available at: https://www.nature.com/articles/d41586-023-02120-8
European Union AI Act, Regulation (EU) 2024/1689. Entered into force 1 August 2024.
Anthropic. “Exploring Model Welfare.” (2025). Available at: https://www.anthropic.com/research/exploring-model-welfare
Singer, P. Animal Liberation (1975; revised 2023). Available at: https://paw.princeton.edu/article/bioethics-professor-peter-singer-renews-his-fight-animal-rights
Stanford University AI Index Report (2025).
“Legal framework for the coexistence of humans and conscious AI.” Frontiers in Artificial Intelligence (2023). Available at: https://www.frontiersin.org/journals/artificial-intelligence/articles/10.3389/frai.2023.1205465/full
“Anthropic rewrites Claude's guiding principles.” Fortune (January 2026). Available at: https://fortune.com/2026/01/21/anthropic-claude-ai-chatbot-new-rules-safety-consciousness/
Council of Europe Framework Convention on AI and Human Rights. Opened for signature 5 September 2024.
Schwitzgebel, E. “Credence-Weighted Robot Rights?” (January 2024). Available at: https://eschwitz.substack.com/p/credence-weighted-robot-rights
“Can a Chatbot be Conscious?” Scientific American (2025). Available at: https://www.scientificamerican.com/article/can-a-chatbot-be-conscious-inside-anthropics-interpretability-research-on/
Tim Green UK-based Systems Theorist & Independent Technology Writer
Tim explores the intersections of artificial intelligence, decentralised cognition, and posthuman ethics. His work, published at smarterarticles.co.uk, challenges dominant narratives of technological progress while proposing interdisciplinary frameworks for collective intelligence and digital stewardship.
His writing has been featured on Ground News and shared by independent researchers across both academic and technological communities.
ORCID: 0009-0002-0156-9795 Email: tim@smarterarticles.co.uk