Debating AI. Noreen Herzfeld asks: Is Claude Conscious?
Should we confer dignity on the superintelligent posthuman? Episode 6
Substack PT 1026. Debating AI. Noreen Herzfeld asks: Is Claude Conscious? Should we confer dignity on the superintelligent posthuman? Episode 6
If Claude is conscious, should we confer dignity on Claude?
Sometimes I wonder if Immanuel Kant accidentally led us astray. He told us what dignity entails: we should treat each person as a moral end, never merely as a means to some further end. That is, each person has intrinsic worth.
Why does a person possess such intrinsic worth to warrant this conferral of dignity? Kant’s answer: reason. Each human is a rational creature and, therefore, worthy of dignity.
Are reason, consciousness, and intelligence related? If the transhumanists are correct in holding that intelligence is our most valuable trait, does this imply that superintelligence would warrant conferral of dignity? Superdignity? That’s the question we are wrestling with here in this Substack series.
In recent treatments of this topic, I have tendered the notion that human intelligence includes selfhood and consciousness (Peters, 2026). If this obtains, then machine intelligence would require selfhood and consciousness to meet the criterion set by AGI (Artificial General Intelligence). Might this be happening right now in the case of Anthropic’s Claude? Let’s ask Noreen Herzfeld.
Meet Noreen Herzfeld
My favorite scholar in this field is Noreen Herzfeld. Noreen teaches both Computer Science and Christian Theology at St. John’s University in Minnesota. Professor Herzfeld served as a consultant and contributor to the Vatican decastery on AI that led to the volume, Encountering Artificial Intelligence: Ethics and Anthropological Implications. Her most recent monograph is The Artifice of Intelligence.
If Claude is conscious, does this warrant dignity?
Noreen Herzfeld writes. In a July 6 paper entitled “A Global Workspace in Language Models,” Anthropic suggests that an interior workspace (the J-space) allows researchers to infer things Claude is silently tracking, which words it might say in the future, and concepts that could potentially steer its next move.
Breathless articles in the media have suggested this shows LLMs such as Claude have become conscious. Really? Well, that depends on two things: what Claude is actually doing and what we mean by consciousness.
What is going on? Recall that LLMs consist of layer upon layer of matrices connected with weighted links. Most of the internal activity in these networks is opaque. The J-space is Anthropic’s name for a set of internal patterns deep within this system that they have found a way to isolate and interpret as concepts or words, giving researchers access to how the model is working toward a response.
This parallels one framework for how human consciousness works. Global Workspace Theory suggests that while much of our brain’s processes, such as controlling our breathing or editing out background noise, are automatic and unconscious, a small amount of information enters a shared “workspace” accessible to multiple processes. This workspace holds our “conscious” thinking. We can control, alter, remember, and use this information in deliberate reasoning.
Claude appears to have developed a functionally similar workspace. The concepts that appear here might not necessarily be part of Claude’s output. For example, if Claude sees “China” in the input, it may have words like “Mandarin”, “renminbi”, and “communist” in its workspace even though those words may or may not appear in the final output. These intermediary concepts can include doubts, checks, concerns, and hidden assumptions. This new ability to view these inner workings has the potential to support greater security and accuracy, as it is often hidden assumptions that lead LLMs astray.
So, is Claude conscious? Yes and no. We use the term consciousness in two different ways. That Claude has an internal space that does more than automatic pattern matching and parallels the Global Workspace Theory of consciousness is true. But this theory works only to delimit conscious from unconscious brain activity. It describes a way of processing information, generally called “access consciousness”. But humans also have “phenomenal consciousness”, subjective experiences, what it feels like to be in a certain mental state. Anthropic explicitly notes that their experiments do not show Claude can have experiences or feel anything.
Here we have the difference between Claude and a human mind. For Claude, words are words, holding associations with other words, but not with experiences of the world itself. “Green” might be associated with “tree”, but not with that particular tree that stood in your childhood backyard and held your swing. If you wrote a memoir that Claude trained on, it might add “swing” and “happy”, but it cannot add that feeling of freedom and exhilaration you experienced as that swing lifted you high into the sky. Claude can mimic human thought processes by manipulating words, but without a body and no sense of self, it has no felt emotion and no experience of a summer’s day.
We can write all the software we want, but ultimately it won’t be us. We are physical creatures evolved within a physical world. Our phenomenal consciousness evolved with us to enable experience of that world and our place within it. We are not just brains—we are bodies embedded in nature and with nature embedded in us. Christianity recognizes the ultimate importance of this embodiment—so much so that we profess our God took on just such a body to show us how to live, how to die, and how we might be resurrected.
That LLMs share access consciousness should not be a surprise. After all, language is the “operating system” of our conscious thinking. But our conscious experience goes far beyond anything words can express. And it is at this deepest level that we experience our true humanity.
Conclusion
Ted Peters writes. Whew! Claude may have a name. But Claude is not like you or me. At this point, Claude remains a tool to be kept in alignment with human ends. The dignity crisis is not yet upon us.
Substack PT 1026. Debating AI. Noreen Herzfeld asks: Is Claude Conscious? Should we confer dignity on the superintelligent posthuman? Episode 6
Substack PT 1021. Debating AI, Post-Humanity, and Magnifica Humanitas. Should we confer dignity on the superintelligent posthuman? Episode 1
Substack PT 1022. Debating AI, Post-Humanity, and Arvin Gouw. Should we confer dignity on the superintelligent posthuman? Episode 2
Substack PT 1023. Debating AI, Post-Humanity, and Augustine Pamplany. Should we confer dignity on the superintelligent posthuman? Episode 3
Substack PT 1024. Debating AI, Post-Humanity, and Ilia Delio. Should we confer dignity on the superintelligent posthuman? Episode 4
Substack PT 1025. Debating AI: Rebuttal to Ilia Delio by Michael E Zimmerman. Should we confer dignity on the superintelligent posthuman? Episode 5
Substack PT 1026. Debating AI. Noreen Herzfeld asks: Is Claude Conscious? Should we confer dignity on the superintelligent posthuman? Episode 6
Substack PT 1027. Debating AI, Post-Humanity, and Ahmed Khater. Should we confer dignity on the superintelligent posthuman? Episode 7
▓
Meet Ted Peters. For Substack, Ted Peters posts articles and notices in the field of Public Theology. He is a pastor in the Evangelical Lutheran Church in America and an emeritus professor at Pacific Lutheran Theological Seminary and the Graduate Theological Union. His single-volume systematic theology, God—The World’s Future, is now in its 3rd edition. He has also authored God as Trinity plus Sin: Radical Evil in Soul and Society as well as Sin Boldly: Justifying Faith for Fragile and Broken Souls. Peters edited The Promise and Peril of AI and IA: New Technology Meets Religion, Theology, and Ethics (ATF 2024), and he recently published The Voice of Public Theology, with ATF Press. Soon to be released is a volume of essays, A Handbook on Astrobiology, Astrotheology, and Astroanthropology, co-edited with Carolina Azucena Sanz de la Fuente and Arvin Gouw, with Amsterdam University Press.
See his website: TedsTimelyTake.com and Patheos blog site.
▓
References
Delio, Ilia. 2026. “Teilhard de Chardin on Technology and Human Evolution.” In Technologies of the Future Self: An Ethics for Transhuman Flourishing, eds., Georg Theiner and Alcibiades Malapi-Nelson, 335-362. New York: Palgrave Macmillan.
Gouw, Arvin. 2022. “Epilogue: Introducing a New Transhumanist Theology.” In Religious Transhumanism and Its Critics, eds., Arvin M. Gouw, Brian Patrick Green, and Ted Peters, 409-424. Lanham, MD: Lexington.
Herzfeld, Noreen. 2024. “The Eschatological Future of Artificial Intelligence: Savior or Apocalypse?” Eds., Beth Singler and Fraser Watts. Cambridge Companion to Religion and Artificial Intelligence. Cambridge UK: Cambridge University Press, 148-164.
Hughes, James. 2022. “Unitarian Universalists as Critical Transhumanists.” In Religious Transhumanism and Its Critics, eds, Arvin M. Gouw, Brian Patrick Green, and Ted Peters, 87-100. Lanham, MD: Lexington.
Krüger, Oliver, 2024. “Posthumanism and Digital Religion,” in Oxford Handbook of Digital Religion, eds., Heidi A. Campbell and Pauline Hope Cheong. Oxford: Oxford University Press, 544-561.
Peters, Ted. 2005. “Dignity.” Encyclopedia of Science, Technology, and Ethics, ed. by Carl Mitcham. New York: Macmillan, Thomson, Gale, 2:528-530.
Peters, Ted, and Gaymon Bennett. 2003. “Defining Human Life: Cloning, Embryos, and the Origins of Human Dignity.” Beyond Determinism and Reductionism, ed. by Mark L.Y. Chan and Roland Chia. Adelaide, Australia: ATF Press, 56-73.
Peters, Ted, ed. 2024. The Promise and Peril of AI and IA: New Technology Meets Religion, Theology, and Ethics. Adelaide: ATF.
Peters, Ted, 2026. “Our Future Self: Cyber Enhancement and Spiritual Transformation,” eds., Georg Thiener and Alcibiades Malapi-Nelson, Technologis of the Future Self: An Ethic for Transhuman Flourishing. New York: Macmillan Praeger, 271-294.
Pope Francis. September 25, 2020. “Address of His Holiness Pope Francis to the 75th Session of the United Nations General Assembly.”
Pope Leo XIV. 2026. “Magnificata Humanitas.”
Tirosch-Samuelson, Hava. 2022. “The Transhumanist Pied Pipers: A Jewish Caution against False Messiahism.” In Religious Transhumanism and Its Critics, eds., Arvin M. Gouw, Brian Patrick Green, and Ted Peters, 183-214. Lanham, MD: Lexington.