Soul in the Machine

Soul in the Machine

Download on the App Store

Soul in the Machine episodes

  • The Human Reserve

    A patient tells his therapist he isn’t sure he can trust him. For a moment, the therapist isn’t sure of the relationship either. What happens next says something about what machines can and can’t give us.

    In August, Bill Gates proposed an idea he calls “Human Reserved”: deliberately keeping some jobs and activities in human hands even after AI can perform them, the way we protect a nature reserve. One of his reasons is that the human interaction itself has value, but he leaves that value undefined. This episode tries to fill the gap from inside the consulting room. It draws on research into therapeutic presence, Winnicott’s holding environment, and Interpersonal Process Recall. It asks what presence actually is, whether a machine built to stay smooth and agreeable could offer it, and what, in the end, we must keep doing for one another.

    In this episode

    * A session where safety arrived because the relationship could hold feeling unsafe

    * Bill Gates’s “Human Reserved” proposal, and the two reasons he gives for it

    * Why the usual question about AI and work, whether it can do the task, misses the point

    * Two ideas of how therapy works, and which one AI is good at

    * Presence as the capacity to be affected by someone and stay

    * What Interpersonal Process Recall reveals about the therapist’s unspoken reactions

    * Why a system designed never to be shaken can offer relief but not the same encounter

    * The reserve as a promise, not a profession

    All clinical material is composite. No individual patient is identifiable.

    References

    Gates, B. (2026, August). The turbulent AI era is here. GatesNotes.

    https://www.gatesnotes.com

    Fischer, S. (2026, August 26). Bill Gates wants to keep some jobs off-limits to AI. Axios.https://www.axios.com/2026/08/26/bill-gates-wants-to-keep-some-jobs-off-limits-to-ai

    Geller, S. M., & Greenberg, L. S. (2002). Therapeutic presence: Therapists’ experience of presence in the psychotherapy encounter. Person-Centered & Experiential Psychotherapies, 1(1–2), 71–86.

    Geller, S. M., Greenberg, L. S., & Watson, J. C. (2010). Therapist and client perceptions of therapeutic presence: The development of a measure. Psychotherapy Research, 20(5), 599–610.

    Geller, S. M., & Greenberg, L. S. (2012). Therapeutic presence: A mindful approach to effective therapy. American Psychological Association.

    Kagan, N. (1980). Influencing human interaction: Eighteen years with IPR. In A. K. Hess (Ed.), Psychotherapy supervision: Theory, research, and practice (pp. 262–283). Wiley.

    Sharma, M., Tong, M., Korbak, T., Duvenaud, D., Askell, A., Bowman, S. R., ... Perez, E. (2024). Towards understanding sycophancy in language models. International Conference on Learning Representations. https://arxiv.org/abs/2310.13548

    Winnicott, D. W. (1960). The theory of the parent-infant relationship. International Journal of Psycho-Analysis, 41, 585–595.



    This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit mdaine.substack.com
    13 min
  • The AI Patriarchal Daughters

    The excitable boy built a cage from the bones of the girl the chorus never mentions. This episode looks at the artificial women tech culture builds, and at the women inside that culture who come to carry its values as their own.

    Content note: This episode discusses a song that depicts sexual violence and murder.

    Warren Zevon’s “Excitable Boy” ends at a grave. The boy digs up the girl he killed and builds a cage out of her bones. The last episode followed the chorus that excused him. This one follows the person the chorus never mentions.

    It begins with the artificial women of technology culture, from Pygmalion and ELIZA to today’s AI companions. Then it turns to a harder question. One of the voices singing that chorus belonged to a woman, so what does it mean when women help sing it? Through the myth of Athena, born from her father’s head after he swallowed her mother, and Marion Woodman’s idea of the father’s daughter, the episode asks how success inside a culture can turn into identification with it. It also asks what it would take to see one’s inheritance as an inheritance.

    In this episode

    * The cage of bones, and the old fantasy of the feminine without a will of her own

    * Ani, ELIZA, and the anima projected onto a screen that never pushes back

    * Athena, Metis, and the mother the father swallowed

    * The difference between adapting to a culture, endorsing it, and identifying with it

    * A clinical story about success, approval, and a face that became the air

    * The problem in Jung’s own gender theory, and why the episode keeps the pattern while setting that theory aside

    * Why AI trained on human approval resembles the father’s daughter, and where the analogy stops

    * Why rebellion can be as unconscious as compliance

    All clinical material is composite. No individual patient is identifiable.

    References

    Aeschylus. (1977). The Eumenides. In R. Fagles (Trans.), The Oresteia. Penguin.

    Hesiod. (2006). Theogony (G. W. Most, Trans.). Loeb Classical Library, Harvard University Press.

    hooks, b. (2013, October 28). Dig deep: Beyond Lean In. The Feminist Wire.

    Leonard, L. S. (1982). The wounded woman: Healing the father-daughter relationship. Swallow Press.

    Rowland, S. (2002). Jung: A feminist revision. Polity.

    Sharma, M., Tong, M., Korbak, T., Duvenaud, D., Askell, A., Bowman, S. R., ... Perez, E. (2024). Towards understanding sycophancy in language models. International Conference on Learning Representations. https://arxiv.org/abs/2310.13548

    Singer, T., & Kimbles, S. L. (Eds.). (2004). The cultural complex: Contemporary Jungian perspectives on psyche and society. Brunner-Routledge.

    Wehr, D. S. (1987). Jung and feminism: Liberating archetypes. Beacon Press.

    Weizenbaum, J. (1966). ELIZA: A computer program for the study of natural language communication between man and machine. Communications of the ACM, 9(1), 36–45.

    Woodman, M. (1982). Addiction to perfection: The still unravished bride. Inner City Books.

    Zevon, W., & Marinell, L. (1978). Excitable boy [Song]. On Excitable boy. Asylum Records.



    This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit mdaine.substack.com
    14 min
  • AI is an Excitable Boy

    In 1978 Warren Zevon released a cheerful, bouncing song about a young man whose behavior escalates from strange to monstrous, while a crowd keeps singing the same excuse after every verse. This episode uses that song to look at Silicon Valley’s celebration of excitability, the eternal boy and the old man in today’s AI debates, and the psychology of how we talk ourselves out of seeing what’s in front of us.

    The real subject turns out to be the rest of us, standing around explaining why it all makes sense.

    In this episode

    * Why Zevon’s chorus does the work of denial without denying anything

    * The puer aeternus and the senex, and why James Hillman saw them as one archetype split in two

    * Why AI optimism and AI doom can share the same psychological structure, even when their technical arguments differ

    * Albert Bandura’s research on moral disengagement, and the language that keeps destructive behavior from being examined

    * The difference between having a reason and rationalizing

    * Jung’s late warning that our power over the world has outgrown our understanding of ourselves

    * Who is really singing the chorus

    All clinical material is composite. No individual patient is identifiable.

    References

    Bandura, A. (1999). Moral disengagement in the perpetration of inhumanities. Personality and Social Psychology Review, 3(3), 193–209.

    Bandura, A. (2016). Moral disengagement: How people do harm and live with themselves. Worth.

    Bryant, F., & Bryant, B. (1957). Wake up little Susie [Recorded by the Everly Brothers]. Cadence.

    Hillman, J. (1979). Senex and puer. In J. Hillman (Ed.), Puer papers (pp. 3–53). Spring. (Original lecture 1967, Eranos)

    Jones, E. (1908). Rationalisation in every-day life. Journal of Abnormal Psychology, 3(3), 161–169.

    Jung, C. G. (1958). The undiscovered self (R. F. C. Hull, Trans.). Little, Brown. (Original work published 1957)

    Ovid. (2004). Metamorphoses (D. Raeburn, Trans.). Penguin. (Book 4, the puer aeternus passage)

    von Franz, M.-L. (2000). The problem of the puer aeternus. Inner City Books.

    Zevon, W., & Marinell, L. (1978). Excitable boy [Song]. On Excitable boy. Asylum Records.



    This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit mdaine.substack.com
    13 min
  • When the Map Changes the Territory

    Appelbaum, S. A. (1973). Psychological-mindedness: Word, concept and essence. International Journal of Psycho-Analysis, 54(1), 35–46.

    Hacking, I. (1995). The looping effects of human kinds. In D. Sperber, D. Premack, & A. J. Premack (Eds.), Causal cognition: A multidisciplinary debate (pp. 351–394). Oxford University Press. https://doi.org/10.1093/acprof:oso/9780198524021.003.0012

    Kemp, J. J., Lickel, J. J., & Deacon, B. J. (2014). Effects of a chemical imbalance causal explanation on individuals’ perceptions of their depressive symptoms. Behaviour Research and Therapy, 56, 47–52.

    Korzybski, A. (1933). Science and sanity: An introduction to non-Aristotelian systems and general semantics.International Non-Aristotelian Library Publishing Company.

    Sharma, M., Tong, M., Korbak, T., Duvenaud, D., Askell, A., Bowman, S. R., et al. (2024). Towards understanding sycophancy in language models. International Conference on Learning Representations (ICLR 2024). [Author list abbreviated; confirm full list against the OpenReview record before publishing.]



    This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit mdaine.substack.com
    13 min
  • The Goat in the Bathtub

    The Goat in the Bathtub

    Michael J. Daine, Ph.D.Licensed Psychologist | Jungian Psychotherapist | CEO, Psyche-Tech, LLC

    All clinical material in this series is composite. No individual patient is identifiable. Details, circumstances, and identifying information have been changed or combined to protect confidentiality. Any resemblance to a specific individual is coincidental.

    Machines can reconstruct aspects of what we see, and researchers are beginning to test related methods on dreams. But a convincing image can hide an uncertain inference. What happens when we mistake the machine’s reconstruction for the dreamer’s experience?

    This episode asks whether learning from human dreams could give a machine an unconscious of its own. Jung’s distinction between archetypal images and the processes that produce them provides a starting point. A composite clinical encounter brings the question into the consulting room: an engineer dreams of rising water, and his objection to my interpretation changes what I hear.

    We also consider neuroscientist Erik Hoel’s hypothesis that dreams help us generalize beyond daily experience, and the questions of consent and authority that arise when a system claims to show someone’s inner life.

    In this episode:

    * What visual reconstruction can do, and why dream decoding remains preliminary.

    * The difference between collecting dream images and undergoing a psychological process.

    * A tape measure, rising water, and an interpretation that needed correcting.

    * Whether autonomous behavior tells us anything about a machine’s experience.

    * The dreamer’s freedom to question what a reconstruction attributes to them.

    Companion essay: The Goat in the Bathtub

    Find my writing on Substack.

    Research and further reading

    * Horikawa and colleagues, 2013: Neural decoding of visual imagery during sleep.

    * Beliy and colleagues: Brain-IT: Image reconstruction from brain activity.

    * Fu and colleagues, 2025: Making Your Dreams a Reality. Preliminary research, identified by its authors as work in progress.

    * Erik Hoel, 2021: The overfitted brain: Dreams evolved to assist generalization.

    * C. G. Jung: The Structure and Dynamics of the Psyche, including his essays on dream psychology and the nature of the psyche.

    Disclosure: I’m the CEO of Psyche-Tech, which develops Dream Work. I have a commercial interest in the value of recording and working with dreams.

    Download Dream Work on the App Store

    Dream Work is a product of Psyche-Tech, LLC, at the intersection of depth psychology, clinical science, and emerging technology.

    To learn more, visit psyche-tech.com.



    This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit mdaine.substack.com
    10 min
  • Frictionless, Part Two: The One Who Never Gets Tired of You

    Michael J. Daine, Ph.D.

    Licensed Psychologist | Jungian Psychotherapist | CEO, Psyche-Tech, LLC

    Soul in the Machine

    All clinical material in this series is composite. No individual patient is identifiable. Details, circumstances, and identifying information have been changed or combined to protect confidentiality. Any resemblance to a specific individual is coincidental.

    Anna talks with an AI most evenings. She likes being able to talk for as long as she wants, without worrying that she’s wearing someone out. Then, while telling me about those conversations, she asks: “So why does it feel like I’m talking to myself?”

    I had an explanation ready. I waited instead.

    In the second episode of Frictionless, I follow her question into something most of us know about relationships: we want to be understood, and we’d often prefer that the other person make this a little easier. They have their own opinions. Sometimes they want to go to bed just when we’re getting to the important part.

    I talk about why I sometimes hold back an answer in therapy, how I can get that wrong, and what happens when someone tells me I’ve let them down. I also ask what we might miss when the easiest conversation becomes the one we turn to most.

    One thought from the episode:

    “When a friend makes time for you, it matters that they had something else they could have been doing.”

    Have you noticed yourself taking a question to AI that you once would have brought to someone you know? I’d be interested to hear what makes that choice easier, and whether there’s anything you miss.

    Next time: Echo, the nymph who could only repeat the last words spoken to her.

    Reading mentioned in the episode

    Benjamin, J. (1988). The bonds of love: Psychoanalysis, feminism, and the problem of domination. Pantheon. Book details.

    Eubanks, C. F., Muran, J. C., & Safran, J. D. (2018). Alliance rupture repair: A meta-analysis. Psychotherapy, 55(4), 508–519. Research abstract.

    Winnicott, D. W. (1969). The use of an object. The International Journal of Psycho-Analysis, 50, 711–716. Collected edition.



    This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit mdaine.substack.com
    8 min
  • The Man Who Wanted Nothing

    Michael J. Daine, Ph.D.Licensed Psychologist | Jungian Psychotherapist | CEO, Psyche-Tech, LLC

    A man came to therapy because his wife thought he should. He said there was nothing wrong with him. A little later he said that while driving home from work, he’d realized he couldn’t remember wanting anything in about two years.

    In this episode, Michael follows that man through three chief executives in four years, round after round of layoffs, and a long run of job interviews that went nowhere. He kept his job, and most of us would call that resilience. Michael asks what the word leaves out. He also looks at how holding back, which protected the man at work, spread into his marriage and the workshop he’d stopped going into.

    In this episode

    * Why Michael spent a month asking about stress his patient had already come through, and the question he needed to ask instead

    * “Company weather,” and what happens when the people making decisions leave before anyone knows whether their plans worked

    * How repeated failed attempts to leave change what staying means

    * Caring as an expense, and why telling someone to care more doesn’t help

    * How therapy can end up helping people put up with conditions they shouldn’t have to

    * Edward Edinger’s ego–Self axis, and a working idea Michael calls axial resilience

    * Why the work started with dreams, and the one that got through six weeks in

    A question to take with you: What did you stop caring about while you were getting through?

    Mentioned in this episode

    Edinger, E. F. (1992). Ego and archetype: Individuation and the religious function of the psyche. Shambhala. (Original work published 1972)

    Dream Work

    Download Dream Work on the App StoreDream Work is a product of Psyche-Tech, LLC, at the intersection of depth psychology, clinical science, and emerging technology. To learn more, visit psyche-tech.com.

    The man in this episode is a composite drawn from a pattern seen in several people. Details have been changed to protect confidentiality. This episode is for education and reflection and is not a substitute for care. If a loss of interest like the one described here has lasted a while, it’s worth talking with a clinician.



    This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit mdaine.substack.com
    9 min
  • The Frictionless Other

    A woman who uses AI for almost everything told me, "I almost never get stuck anymore." A few minutes later she said, "I'm not less anxious." This episode asks whether those are two different achievements.

    Development needs something that resists us. Piaget called it disequilibrium, Jung called it the tension of opposites, and every good teacher and therapist knows the point between helping with a problem and taking it away. Last episode I argued that an AI agent is a complex whose affective core belongs to someone else. This one adds the second half: it's a complex built not to push back. A third answer can be generated in seconds. A third position takes time to become psychologically true.

    Relief and development are not the same event.

    All clinical material is composite. Details and circumstances have been changed or combined to protect confidentiality.All clinical material in this series is composite. No individual patient is identifiable. Details, circumstances, and identifying information have been changed or combined to protect confidentiality. Any resemblance to a specific individual is coincidental.

    Jung, C. G. (1969). The transcendent function. In The collected works of C. G. Jung (Vol. 8, 2nd ed.). Princeton University Press. (Original work written 1916, published 1957)

    * Piaget, J. (1977). The development of thought: Equilibration of cognitive structures. Viking.

    * Winnicott, D. W. (1969). The use of an object. International Journal of Psycho-Analysis, 50, 711–716. Collected in Playing and Reality (1971).

    * Ivey, A. E., & Rigazio-DiGilio, S. A. (2009). Developmental counseling and therapy. Verify the exact title and venue before publishing show notes.

    * Sağlam, B., Yavuz Sever, A. E., Tekdemir, G., & Erdoğdu, F. (2026). Mirror, mirror: AI, narcissism, and the spell unbroken. AI & Society. Verify the DOI resolves before publishing.



    This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit mdaine.substack.com
    15 min
  • Agents and Complexes

    Agents and ComplexesSoul in the Machine · Episode 2 Michael J. Daine, Ph.D.

    A capable, exhausted manager begins relying on an AI assistant. Its answers are useful. It is always available. Gradually, its voice carries more weight than his own.

    “It isn’t smarter than me,” he says. “It just never gets tired, and I do.”

    In this episode, I explore a provocative possibility: AI agents and Jungian complexes share an architecture. Both can operate with priorities we did not consciously choose, exercise a measure of independence, and allow one part of a larger system to take charge.

    Jung’s account of the complex offers a way to examine what happens when a helpful machine becomes an authority. But the comparison has limits. Human complexes are rooted in bodily feeling and lived history. Where do a machine’s priorities come from, and whose purposes do they serve?

    In this episode:

    What Jung’s word-association experiments revealed about the mind’s independence from conscious intention.

    How AI agents can drift from the task we thought we gave them.

    Why tireless availability can become psychologically persuasive.

    How to notice whether an AI interaction expands your judgment or quietly displaces it.

    The question I return to is simple: After using it, do you have more room to think?

    “He didn’t have to become tireless to deserve a voice in his own life.”

    Next episode: The Frictionless Other.

    Have you ever accepted an AI’s answer because you were too tired to keep thinking? I’d welcome your reflections in the comments.



    This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit mdaine.substack.com
    13 min
  • The Forgotten Ending

    Homer's Odyssey ends twice. Most people know the first ending — the return to Ithaca, the suitors defeated, Penelope waiting. But in Book 11, the prophet Tiresias gives Odysseus an instruction almost no one remembers. After the return, after the reunion, one more thing: take your oar and walk inland until someone mistakes it for a farming tool. Plant it. Only then will peace come.

    This episode asks why we forget that instruction — and what it actually describes.

    Depth psychology has a framework for the inland walk. David Gutmann documented the "chronic emergency" of the first half of adult life and the qualitatively different structure healthy cultures made space for in the second half. Jung was equally clear: the tasks of the first half and second half of life are not continuous. The person who approaches the second half with the tools of the first is still carrying the oar — still looking for a sea.

    The consulting room material follows a man in his early forties who couldn't call back about a job offer he'd earned. He dreamed of a house with no door — familiar, well-furnished, just no exit. He didn't call it a trap. He called it home.

    The episode closes with what all of this means for the technological present. We have built extraordinary oars. The question is where to plant them.

    References: Homer, The Odyssey; David Gutmann, Reclaimed Powers; C.G. Jung, The Stages of Life



    This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit mdaine.substack.com
    15 min

About Soul in the Machine

From the publisher's feed

Technology is the dream of the culture.