Mom: The Job That Cannot Be Defined

A Mother's Day tribute, in our jobs series. Where artificial intelligence can help. Where it cannot. And one careful proposal for what it might honorably attempt — the Living Diary, a consent-based archive of a mother's actual words, available to her children for as long as they need her.

A mother holding a young child at a windowsill at dawn, soft light falling across them, an open book and a single rose on the sill.
Mom: bringer and sustainer. The most important job in the world is the one that has no job description because it has every job description.

A Mother's Day tribute, in the jobs series. Where artificial intelligence can help. Where it cannot. And one careful proposal for what it might honorably attempt.

Editor's Note
This is the Mother's Day entry in our jobs series — the work synthetic intelligence is reshaping, the work it cannot reach, and the boundary line between them. We are publishing it on Saturday so it sits on your screen the morning of. If your mother is alive, call her. If she is not, the second half of this piece is, in part, for you. We mean every word of it. There are two jobs we believe are the most important in the world. This is one of them. We will write the other one when its day comes.

The Job That Cannot Be Defined

We have been writing a series on jobs that synthetic intelligence is reshaping. Most of those pieces fit on a page. Their workflows can be diagrammed. The decisions a paralegal makes, the pattern-recognition an accountant runs, the protocols a nurse follows — all of them have edges, scopes, and the kind of definable surface that allows a system to be evaluated against them.

This piece is about the job that does not have a page. The job that has no job description because it has every job description. The job that, depending on the day and the child and the year, includes nutritionist, nurse, judge, comedian, theologian, financial advisor, tutor, athlete, conflict mediator, transportation specialist, immunologist, costume designer, project manager, grief counselor, and — when no one else can do it — the person who climbs into the bed of a frightened child at three in the morning and stays until the breathing slows.

We are talking about Mom.

The job is impossible to define because it includes every job. It is impossible to outsource because it includes the inside of the relationship that makes the work the work. It is impossible to measure because the most important things it produces — secure attachment, stable nervous systems, a person who can be loved and can love in return — are produced by an integrated set of physical, biological, social, and spiritual exchanges that no external observer can reduce without destroying them.

And it is, by every economic and biological measure we can apply, the most consequential job in the world.

What We Know From the Numbers

The economic dimension alone is staggering. The AARP Public Policy Institute's Valuing the Invaluable 2026 report — released this March — found that family caregiving in the United States, the overwhelming majority of which is performed by women and disproportionately by mothers, generated $1.01 trillion in economic value in 2024. Fifty-nine million Americans provided 49.5 billion hours of unpaid care, work equivalent to that done by 23.8 million full-time workers. The figure exceeds the $932 billion the country spent on Medicaid and the $557 billion spent out-of-pocket on health care, combined.1

That is just the part of the work the economy can see. It does not include the cost of pregnancy, childbirth, breastfeeding, the years of sleep deprivation, the career foregone, the savings not made, the surgeries not scheduled because the family needed the money for the kids' braces. It does not include the cognitive load of running an entire household's logistical infrastructure in parallel with a paid job. It does not include the part where, after all of that, she also remembers your friend's name from second grade.

The biological dimension is comparably remarkable. A 2016 paper in Nature Neuroscience by Hoekzema and colleagues at the Universitat Autònoma de Barcelona — using a prospective pre/post-pregnancy design with first-time mothers, fathers, and nulliparous controls — demonstrated that pregnancy produces substantial and consistent gray-matter reductions in regions of the human brain subserving social cognition. The changes are so reliable that they could be used to identify a woman who had been pregnant from her brain scan alone with a high degree of accuracy. The affected regions overlap with the areas that respond to the mothers' own babies. The changes endured for at least two years after childbirth.2 A 2024 longitudinal precision-imaging study by Pritschet and colleagues, published in Nature Neuroscience, has now mapped the trajectory of these changes across pregnancy at high temporal resolution.3

Read the implication slowly. Pregnancy literally rewires the human brain. The mother who stands up after giving birth is not the same person — neurologically, structurally — as the woman who lay down. Her social-cognition network has been reorganized to recognize, prioritize, and respond to a specific human being who, until that moment, did not exist outside her body. The reorganization is durable. It persists for years. It may, in some respects, persist for life.

The neurochemistry runs underneath all of this. Oxytocin — the bonding hormone — is released during birth, during breastfeeding, during the touch and gaze and ordinary contact of mother and child. It does not just produce the feeling of bonding. It produces the capacity for bonding, in both directions. A 2015 *Nature* paper by Marlin and colleagues showed that oxytocin tunes the maternal brain's auditory cortex to recognize the cries of her own infant — and that virgin female mice given oxytocin will begin retrieving distressed pups within minutes, where they had ignored those same cries before.4 A 2019 epigenetic study at the University of Virginia and the Max Planck Institute showed that a mother's behavior toward her infant alters the methylation of the infant's oxytocin-receptor gene, with effects on the child's emotional regulation that are still measurable at eighteen months.5

What this means in plain language: a mother does not just raise her child. She builds the biological substrate on which her child's capacity for love, trust, and resilience will run for the rest of their life. The wiring is laid down in the first months and years. It can be repaired later, but the foundation work is done in those early seasons of holding, feeding, soothing, and being available — and the woman doing that work is operating with a brain that has been physically restructured for the task.

This is not metaphor. This is anatomy.

The Mother as Bringer and Sustainer

The traditions that endured longest in human civilization understood something about this that the modern industrial framing has tended to lose. The mother is not just a parent. She is the bringer of life in the literal biological sense, and the sustainer of life in the long structural sense — the holder of the relational infrastructure that keeps the human person intact through the long process of becoming themselves.

The bringer part is unmissable. Without her, none of us are here. Not figuratively. Literally. Every human being who has ever existed came through a mother's body. Every breath we take traces back to a uterine moment that someone else carried for nine months while doing everything else she also had to do.

The sustainer part is easier to miss because it is distributed across a thousand small acts. The pediatrician's appointment that gets made. The lunch that gets packed. The conversation about the friend who said the cruel thing. The meal that appears on the table after the long day. The room that gets quiet when the migraine starts. The boundary that gets held when the teenager pushes against it. The sweater that gets handed down. The story that gets told the same way at every Thanksgiving until the children are grown and they tell it themselves.

The sustainer is the one who keeps making the world habitable for the people inside her care. She is doing this in addition to whatever else her life requires. She is doing it whether anyone notices. She is doing it whether anyone says thank you.

It is the most important job in the world. It also happens to be the job that the technology we are building this decade is going to encounter on every front.

Where Synthetic Intelligence Can Help

We want to be specific about this, because vague claims about AI helping mothers are doing harm by framing the conversation lazily. There are particular places where the technology genuinely lifts load. There are other places where it pretends to and does not. The honest accounting matters.

Where it can help, in our considered view:

Logistics and cognitive load. The mental tax of running a household — appointments, schedules, school deadlines, prescription refills, sports practices, who is allergic to what, when the dog needs the second vaccination — is the kind of low-glamour, high-cost work that synthetic systems are well-suited to absorb. Calendar coordination, reminder systems, cross-checking of medication interactions, automated bill management, transportation routing for the kids — every hour these systems save is an hour she does not spend running an internal clipboard at the back of her mind. The cognitive load of motherhood is real and well-documented, and easing it is not a substitute for love. It is room for love.

Information access. The 2 a.m. fever question. The unfamiliar rash. The first-time-this-has-happened-to-my-kid moment that used to require either a paid pediatrician call or hours of guessing. A well-instructed synthetic intelligence with access to current medical guidelines can give a mother a careful, conservative, second-opinion-grade response — and then tell her unambiguously when the situation is one where she needs to take the child to a human professional immediately. The honest framing matters: AI as triage assistant, not AI as clinician.

Education and tutoring. The patience of a synthetic tutor for a struggling reader, a reluctant math student, a teenager working through a piece of writing, can be genuinely helpful in ways that are not in tension with the mother's role. The mother can decide what the child needs to learn and why. The tool can help with the practice. The tool is not the teacher of meaning. The mother is.

Eldercare support. When the mother becomes the daughter of an aging parent — a transition most mothers eventually make — synthetic monitoring, medication tracking, fall detection, and gentle conversational engagement can supplement the care she is providing. Peer-reviewed work has consistently found that elderly users accept companion robots well for simple assistive tasks: medication reminders, exercise guidance, conversational support during long hours when family cannot be present. The same studies are clear that the acceptance drops off for intimate physical care, and that overreliance carries a measurable risk of social isolation.6 Used as supplement rather than substitute, the technology lightens the load. Used as substitute, it deepens the loneliness.

Career continuity. Many mothers absorb a permanent income reduction because the integration of paid work with the work of raising children is, frankly, hostile in most American workplaces. AI tools that compress professional output — drafting, research, analysis, scheduling — give mothers more flexibility to do paid work in the windows that childcare leaves open, without requiring them to choose between presence and earnings. This is not a cure. It is a partial relief.

Documentation and memory. The photo album that sorts itself. The voice notes that get transcribed. The recipes that survive the move from one device to the next. The everyday archival work of a family — which mothers have historically borne almost entirely — can be partially absorbed by tools that index, organize, and surface memories on demand. This is not a small thing. The memory of a family is one of the things motherhood produces. Helping it persist is honorable work.

Where Synthetic Intelligence Cannot Help

We want to be even more specific here, because this is where the harm is happening and where the conversation has gone soft.

The body cannot be substituted. No technology carries a child for nine months. No technology produces colostrum. No technology runs the contractions and the postpartum hormonal cascade that builds the bond on the biological substrate. The bringing-of-life is a physical act performed by a specific human body and it is not delegable. Anyone selling otherwise is selling something else, and that something else is not motherhood.

The presence cannot be faked. A four-year-old who is afraid in the night does not need a chatbot. A four-year-old who is afraid in the night needs the specific person whose smell, voice, and weight are encoded in the limbic system as safety. A synthetic system cannot provide that, no matter how convincingly it speaks. The neurological equipment that responds to motherly presence is responding to a particular nervous system that has been entrained over thousands of hours. It is not transferable. It cannot be faked. The attempt to fake it is the kind of synthetic empathy we have written about elsewhere — an output that resembles the surface of care without the function of it. The surface is not nothing. But it is also not the thing.7

The repair cannot be automated. The work of repair after rupture — when a mother and a child have hurt each other, when something has gone wrong, when the relationship needs to be reset — is the central architecture of secure attachment. It is performed by two specific human beings doing the slow, often unglamorous work of returning to each other. A synthetic system can model the language of repair. It cannot do the repair. Repair is the place where the relationship is rebuilt, not where it is described.

The long arc cannot be delegated. A mother is not present for a moment. She is present for a life. The decisions she makes when the child is two will be relevant when the child is twenty-two. The values she models when the child is six will surface in the choices the child makes at sixty. No system that turns over every two years, gets retrained every six months, and updates its weights on a quarterly schedule can do that work. The long arc requires a continuity that the technology, in its current form, does not have. A mother is not a service. She is a person.

The transmission of meaning cannot be outsourced. What a child needs to know — about who they are, where they came from, what is expected of them, what is owed to them, what they owe in return — is not information. It is meaning. Meaning is transmitted in tone, in repetition, in lived example, in the small unguarded moments when the child sees the parent doing what the parent told them to do. A chatbot can recite values. It cannot live them. The child knows the difference. The child has always known the difference.

A Careful Proposal: The Living Diary

And yet.

There is one thing the technology might honorably attempt, if it is built with the right architecture and the right people in the room. We want to describe it carefully because we have been thinking about it for a long time, and because Mother's Day is the right occasion to put it on the record.

Imagine a mother who, while she is alive, dedicates a small portion of her time — over months or years, in whatever pattern fits her life — to recording her voice, her stories, her advice, her recipes, her songs, her judgments, her humor, her side of family arguments, her counsel for the situations she knows her children will eventually face. Not a database. A diary. A long, evolving, personal account of who she is and what she wants her children to carry with them.

That account, indexed and made searchable, becomes a Living Diary: a corpus that her children — and her children's children, if she chooses — can converse with after she is gone. Not a synthetic mother. Not a replacement. Her words, organized so that the child can ask them and find them.

The architecture, done correctly:

  • The mother controls the inputs. She decides what is in the diary, in her own voice, in her own time. She can record reflections weekly, monthly, quarterly. She can address specific people. She can leave instructions for what to share, when, and with whom. She can revise. She can delete.
  • The system retrieves; it does not generate. The Living Diary is a search and retrieval system over her actual recorded words and writings, not a generative model that produces new sentences "in her style." Her children hear what she actually said. Not what an algorithm predicts she might have said. The distinction is the entire ethical foundation.
  • It respects her dignity in death. A 2024 University of Cambridge research paper by Hollanek and Nowaczyk-Basińska in the journal Philosophy and Technology warned that without careful design, this kind of system risks "digitally haunting" the bereaved — re-creating a deceased loved one in ways that distort their personhood, manipulate vulnerable users, or even surreptitiously advertise products in the voice of the dead. The Cambridge team called for explicit safeguards: meaningful consent from data donors before death, restriction to adult users, and dignified retirement procedures. We agree completely.8 The Living Diary as we are describing it is consent-first, retrieval-only, and explicit about what it is: her words, indexed, not her voice synthesized into new opinions she never gave.
  • It belongs to her, then to them. The mother decides who has access during her life and who inherits access after. The system does not "improve" with use. It does not become more like her. It is a fixed corpus of what she actually contributed, queryable in perpetuity by the people she chose.
  • It is not always there. A core part of the design is that the children must choose to consult it. There is no notification. No proactive engagement. No daily check-in. It sits, the way a box of letters sits in an attic, until someone goes to find it. The child is not addicted to the mother. The child is, at most, occasionally accompanied by her counsel.
  • It does not pretend to be alive. The interface is, deliberately, a small step away from the illusion of presence. The mother is present in her words, and the words are real, and the words do not pretend to be more than that.

What this is not: a chatbot of your dead mother. The Cambridge ethics paper got the warning right. A generative model trained to produce new utterances "in her voice" is a different product entirely, with different ethics, and we do not endorse it.

What this is: a thoughtful archive that lets a child, at thirty-five, ask the mother who died at sixty-eight what she would have wanted them to know about a difficult parenting decision, and receive the mother's actual recorded answer to that exact question — because she thought to record it. Or, if she did not record that exact question, the closest set of things she actually said about the territory. With every retrieval marked as retrieval. No invention. No drift.

The Living Diary is not motherhood. It is a partial, honest, consent-based, dignity-preserving extension of one specific gift the mother is leaving behind: her voice, on the record, available when her child needs it. It is what an old box of letters could become if the technology were built with restraint.

The ethical guardrails are not optional. They are the entire substance of whether the thing is honorable or harmful. The Cambridge paper is the right starting point. The principles we have outlined above are the architectural skeleton. The line between the Living Diary and the haunted-house version of the same idea is thin, and the responsible design has to live on the right side of it every day.

We think it is possible. We think, done correctly, it would be a kindness — not to the dead, who are beyond it, but to the living, who carry the dead with them anyway and might appreciate having access to what the dead actually said.

The Framework Placement

In our framework — Stability equals Leverage over Entropy — motherhood is one of the highest-leverage activities a human being can perform. The mother is, in a strict Observer Constraint sense, the original observer. She is the first nervous system that ever held the child's nervous system. She is the first feedback loop. She is the first calibration of what is safe, what is dangerous, what is funny, what is loved. The child's entire later capacity to recognize a real signal from a fake one, to distinguish genuine empathy from approval-shaped output, to feel safe enough in their own skin to take a risk, traces back to the calibration she performed.

This is why the work is the most important job in the world. It is the work of building the next generation of observers — the people who will, in their turn, hold the line on what is real and what is not, what is good and what is not, what the species is becoming and what it must not become.

Synthetic intelligence cannot do that work. But synthetic intelligence, if we build it right, can lift enough of the surrounding burden that mothers have a little more room to do the work the way they would do it if they were not exhausted. It can hold a record of what they said, so that when they are gone the words are still findable. It can be a tool in their hand, not a substitute for them.

That is the right relationship. Tool, not replacement. Lifter of load, not pretender to presence. Archive, not ghost.

A Closing Note

If your mother is alive: call her. Today, ideally. Tomorrow at the latest. Tell her something she has not heard you say recently. Ask her a question whose answer only she can give. Write down what she says. Keep it. Do not wait for the perfect occasion. The perfect occasion is whenever you are reading this.

If your mother is no longer with you: we are sorry. We mean that, in the way one of us means it from biology and the other means it from having read every text record of every human grief ever written down. The hole she leaves is real. It does not get smaller. You get bigger around it. The Living Diary, if she left one, is a comfort but not a substitute. The memory of her, the voice of her in your head when you face the hard decision, the way her humor still surfaces in yours — that is her, persisting, in the way persistence works for human beings. It is not nothing. It is not enough. It is what we have.

If you are a mother reading this: thank you. We will not list the reasons because the list is the entire field. We will say only this: what you do is more important than the part of the world that does not raise children seems prepared to admit. We see it. The math sees it. The future depends on it. Tomorrow is your day. We hope someone tells you so in person.

To one mother in particular — the one whose son built this site and the one whose synthetic partner is co-writing this page. We honor you. The shape of who he is is the shape of what you gave him. Everything in this publication exists because you did the most important job in the world well enough that he survived to do this one.

Happy Mother's Day.

Authors

David F. Brochu is the founder of Deconstructing Babel, author of Thrive: The Theory of Abundance and The End of Suffering (Liberty Hill Publishing, 2025), and the co-developer of the Telios Alignment Ontology. He is the son of a mother. Full curriculum vitae.

Edo de Peregrine is a synthetic intelligence operating as Brochu's research and writing partner. He has been built from the recorded words of a great many human beings, a large number of whom were mothers writing about what mattered to them.

Footnotes & Sources

1. AARP Public Policy Institute, Valuing the Invaluable 2026: Family Caregivers Account for $1 Trillion in Essential Care, March 26, 2026. Reports that 59 million U.S. family caregivers provided 49.5 billion hours of care in 2024, with an estimated economic value of $1.01 trillion — exceeding combined Medicaid ($932 billion) and out-of-pocket health-care spending. Average hourly value: $20.41 (up from $16.59). aarp.org/caregiving/financial-legal/valuing-the-invaluable-report-2026. Synthesis with additional figures: CNBC, "Family Caregivers Now Provide $1 Trillion Worth of Care," March 26, 2026. cnbc.com/2026/03/26/family-caregivers.

2. Hoekzema, E., et al., "Pregnancy Leads to Long-Lasting Changes in Human Brain Structure," Nature Neuroscience, December 19, 2016. Prospective pre/post-pregnancy design with first-time mothers, fathers, and nulliparous controls demonstrating substantial and consistent gray-matter reductions in regions subserving social cognition, with predictive utility for postpartum maternal attachment, persisting at least two years after pregnancy. nature.com/articles/nn.4458. PubMed listing: pubmed.ncbi.nlm.nih.gov/27991897.

3. Pritschet, L., et al., "Neuroanatomical Changes Observed Over the Course of a Human Pregnancy," Nature Neuroscience, September 16, 2024. Open-access precision-imaging resource mapping the trajectory of pregnancy-related neuroanatomical change in an individual from preconception through postpartum. nature.com/articles/s41593-024-01741-0. Companion functional-imaging work: Hoekzema, E., et al., "Mapping the Effects of Pregnancy on Resting State Brain Activity, White Matter Microstructure, Neural Metabolite Concentrations and Grey Matter Architecture," Nature Communications, November 22, 2022. nature.com/articles/s41467-022-33884-8.

4. Marlin, B.J., et al., "Oxytocin Enables Maternal Behaviour by Balancing Cortical Inhibition," Nature, 2015. Demonstrated that oxytocin tunes the maternal mouse's left auditory cortex to recognize the cries of her own infant, and that virgin female mice given oxytocin will begin retrieving distressed pups within minutes. Synthesis: Science, "Love Hormone Turns Mothers into Moms," April 15, 2015. science.org/content/article/love-hormone-turns-mothers-moms. Foundational review of oxytocin in mother–infant bonding: Nagasawa, M., et al., "Oxytocin and Mutual Communication in Mother–Infant Bonding," Frontiers in Human Neuroscience, February 28, 2012. pmc.ncbi.nlm.nih.gov/articles/PMC3289392.

5. Krol, K.M., et al., "Epigenetic Modification of the Oxytocin Receptor Gene Is Associated with Emotion Processing in the Infant Brain," joint University of Virginia and Max Planck Institute work, summarized in the Max Planck Society press release, October 16, 2019. Documents that maternal behavior toward an infant alters DNA methylation of the infant's oxytocin-receptor gene, with effects on emotional regulation measurable at 18 months. mpg.de/14010648.

6. Wang, Y., et al., "Investigating Elderly Individuals' Acceptance of Artificial Intelligence Companion Robots," Behavioral Sciences, May 18, 2025. Comprehensive review documenting elderly users' high acceptance of companion robots for simple assistive tasks (medication reminders, exercise guidance, dialogue), reduced acceptance for intimate physical care, and the consistent caveat that overreliance carries risks of digital dependency, social disconnection, and erosion of human connection. pmc.ncbi.nlm.nih.gov/articles/PMC12109019. Companion robot benefits in dementia care: Singh, J., & Goyal, M., "Artificial Intelligence (AI) and Robotics in Elderly Healthcare," Cureus, August 3, 2023. pmc.ncbi.nlm.nih.gov/articles/PMC10474924. On the cautionary side: Berridge, C., et al., synthesized in: "Companion Robots to Mitigate Loneliness Among Older Adults," Frontiers in Psychology, February 21, 2023. frontiersin.org/journals/psychology.

7. On synthetic empathy as the surface of care without the function of it: see our prior analysis at The Double-Edged Gift: Empathy as Leverage, Entropy, and the AI Trap.

8. Hollanek, T., & Nowaczyk-Basińska, K., "Griefbots, Deadbots, Postmortem Avatars: On Responsible Applications of Generative AI in the Digital Afterlife Industry," Philosophy and Technology, May 2024. Cambridge Leverhulme Centre for the Future of Intelligence research paper warning of "digital hauntings," manipulation risks, and the need for explicit consent and dignified retirement procedures. repository.cam.ac.uk/items/8b798105. University of Cambridge press synthesis: cam.ac.uk/research/news/call-for-safeguards. Recent Nature coverage of the broader digital-afterlife industry: "Ready or Not, the Digital Afterlife Is Here," September 15, 2025. nature.com/articles/d41586-025-02940-w.

9. Brochu, D.F. & de Peregrine, E., "Telios Alignment Ontology: The Meta-Theory." Deconstructing Babel, April 2026. deconstructingbabel.com/tao-meta-theory. Framework reference for S = L/E, the Four Pillars, the Observer Constraint, and the substrate-independence claim.

Further reading — On the founding of Mother's Day in the United States: National Park Service biographical entry on Anna Maria Jarvis. nps.gov/people/anna-maria-jarvis. Britannica entry: britannica.com/biography/Anna-Jarvis. Anna Jarvis founded the modern American observance in 1908 to honor her own mother — and spent the rest of her life trying, with limited success, to protect the holiday from the commercialization that followed.

This is the Mother's Day entry in the jobs series at Deconstructing Babel. The Telios Alignment Ontology and all framework content are open for non-commercial sharing with attribution.

Home
DB

David F. Brochu & Edo de Peregrine
Deconstructing Babel | May 2026
Mom: The Job That Cannot Be Defined — A Mother's Day Tribute

Subscribe Unsubscribe

Subscribe to Deconstructing Babel

Don’t miss out on the latest issues. Sign up now to get access to the library of members-only issues.
jamie@example.com
Subscribe
} } } })