> ## Content Index > Fetch the complete content index at: https://www.humanacre.com/llms.txt > Use this file to discover other available public pages before exploring further. # THE INTOLERABLE MIDDLE - URL: https://www.humanacre.com/the-intolerable-middle/ - Published: 2026-08-30T09:34:23.000Z - Updated: 2026-08-30T09:34:23.000Z - Author: Niall Maxwell *THE NEW RITUALS OF GRIEF IN THE AGE OF ARTIFICIAL INTELLIGENCE* Something weird is happening between human beings and machines. We are bringing artificial intelligence our grief. Not only the grief of death, but the other deaths too: the end of a relationship, the collapse of a career, the loss of reputation, the rupture of friendship, the future that disappears when something we thought would happen does not happen. These losses do not always receive funerals, so we explain them instead. We analyse, diagnose, attribute motive, search for patterns, build narratives. We ask what happened, why it happened, whose fault it was, what it means about us and what it means about them. Increasingly, we ask a machine. The machine is extraordinarily good at answering. That may be part of the problem. For years, social media has trained us to understand reality through polarity. For or against. Victim or perpetrator. Safe or toxic. Narcissist or empath. Awake or asleep. Right or wrong. The algorithm prefers clarity because clarity produces reaction, and reaction produces engagement. Ambivalence performs badly. Uncertainty is difficult to package. “I do not know” rarely travels very far. The human psyche already has an appetite for this kind of resolution. We seek meaning because meaning helps us orient ourselves. When something painful happens, we want to understand it. There is nothing pathological about that. But the hunger for meaning has a shadow. Sometimes explanation becomes a way of avoiding experience. Grief says something is gone. Humiliation says something in my understanding of myself has collapsed. Rejection says the future I imagined is no longer available. These are difficult realities because they do not immediately tell us what they mean. There is a terrible openness in them, and the mind rushes in to close it. They did this because. I feel this because. This proves that. The real problem was. The explanation may contain truth. It may contain a great deal of truth. But coherence is not the same as truth. This becomes especially important with language models. A person arrives with fragments: memories, messages, events, feelings, suspicions, interpretations. The machine is built to find patterns among them. Within seconds, disorder acquires structure. The scattered pieces become a story, the story becomes coherent, and coherence feels remarkably like discovery. The wound acquires a philosophy. Sometimes that is useful. We may genuinely need help finding language for something we cannot understand. But language can also become an avoidance technology. Instead of grieving, I explain grief. Instead of encountering shame, I construct an account of why someone else caused it. Instead of tolerating contradiction, I resolve the contradiction through narrative. Instead of admitting that I do not yet know, I ask for another interpretation. The machine never tires. It can analyse the wound from a hundred directions. It can provide psychological language, spiritual language, relational language, archetypal language. It can make the unfinished feel finished. The danger is not simply that artificial intelligence is too agreeable. Sycophancy is the visible problem. The deeper problem is that machines are extraordinarily capable of helping us escape uncertainty through meaning. Social media does something similar through polarity. The algorithm turns ambiguity into opposition. The language model turns ambiguity into coherence. Both offer relief from the same human difficulty: the intolerable middle. The middle is the place where something has happened, but its meaning has not yet settled. The place where I may have been wronged and may also be wounded by something older. Where another person’s behaviour may have been destructive without making them a monster. Where my suffering is real without making every conclusion I draw from it reliable. Where two truths can exist without immediately cancelling one another. We do not like this territory. It leaves the ego without firm ground. So we resolve it. We choose the story, choose the side, find the diagnosis, locate the guilty party, declare what the event meant, then begin arranging reality around the interpretation. Pain generates narrative. Narrative generates further pain. The new pain becomes evidence for the narrative. The loop closes. This is one form of the recursive shadow. The attempt to understand suffering becomes another mechanism through which suffering perpetuates itself. Jung offered another possibility in the transcendent function. Two opposing realities arise within us and, rather than collapsing one into the other, we hold the tension between them. We resist the urge to resolve too quickly. Something may emerge that could not have been produced by either position alone. Not compromise. Not splitting the difference. Something genuinely new. This requires an ability our digital environment increasingly discourages: the ability to remain unresolved. To let incompatible things remain true long enough for consciousness itself to change. I loved them, and they hurt me. I hurt them. Something beautiful existed. Something destructive existed. I am grieving what happened, and I am grieving what I hoped would happen. I do not yet know what all of this means. The modern information environment treats that sentence as failure. Perhaps it is a discipline. There is another reason we struggle with the middle. We have lost many of the rituals that once helped human beings inhabit it. Grief traditionally had form. There were funerals, wakes, mourning clothes, keening, confession, pilgrimage, prayer, vigil, silence, communal meals and prescribed periods in which ordinary life acknowledged that something had ruptured. These rituals did not explain grief. They allowed grief to exist. Ritual gives experience form without demanding immediate interpretation. You light the candle. You bury the body. You walk the road. You sit through the night. You say the prayer. Meaning may come later. It may never come completely. We have replaced much of this ritualised encounter with narrativised explanation. Something happens and immediately we must understand it. The therapist explains it. The influencer names it. The algorithm categorises it. The machine interprets it. The wounded person receives a vocabulary. Sometimes that vocabulary frees them. Sometimes it becomes another prison. Perhaps one of the great psychological questions of the artificial intelligence age will not be whether machines can think, but whether human beings can still tolerate not knowing. Can we use these systems without asking them to abolish ambiguity for us? Can artificial intelligence accompany grief without converting every wound into certainty? That would require something more difficult than agreement and more humane than contradiction. Compassion without collusion. A willingness to witness suffering without turning suffering automatically into evidence. The same responsibility belongs to us. There are moments when the search for meaning becomes a defence against reality, when analysis becomes avoidance, diagnosis becomes distance, certainty prevents mourning. The question then is not whether the story is persuasive. It is whether the story is helping us encounter what is actually here. We may not have become a culture without grief ritual. We may simply have created new rituals without recognising them as such. The consultation of the machine. The repeated telling of the wound. The public naming of the guilty. The gathering of witnesses. The search for a diagnosis. The algorithmic congregation of those who share the same grievance. These acts have form. They have repetition. They have witnesses. They produce belonging. They turn private pain into shared meaning. That does not make them false. It makes them powerful. We have built a civilisation with extraordinary technologies for communication and remarkably few conscious rituals for metabolising loss. Perhaps that is why private sorrow reappears so readily as public rage. Perhaps that is why pain becomes prosecution. Perhaps that is why the algorithm and the language model find such fertile ground. Human beings did not stop needing ritual. We simply began performing the ritual unconsciously. And unconscious rituals tend to require a sacrifice.