The Last Human Job: The Work of Connecting in a Disconnected World

by Allison J. Pugh

A timely and urgent argument for preserving the work that connects us in the age of automation

With the rapid development of artificial intelligence and labor-saving technologies like self-checkouts and automated factories, the future of work has never been more uncertain, and even jobs requiring high levels of human interaction are no longer safe. The Last Human Job explores the human connections that underlie our work, arguing that what people do for each other in these settings is valuable and worth preserving.

Drawing on in-depth interviews and observations with people in a broad range of professions—from physicians, teachers, and coaches to chaplains, therapists, caregivers, and hairdressers—Allison Pugh develops the concept of “connective labor,” a kind of work that relies on empathy, the spontaneity of human contact, and a mutual recognition of each other’s humanity. The threats to connective labor are not only those posed by advances in AI or apps; Pugh demonstrates how profit-driven campaigns imposing industrial logic shrink the time for workers to connect, enforce new priorities of data and metrics, and introduce standardized practices that hinder our ability to truly see each other. She concludes with profiles of organizations where connective labor thrives, offering practical steps for building a social architecture that works.

Vividly illustrating how connective labor enriches the lives of individuals and binds our communities together, The Last Human Job is a compelling argument for us to recognize, value, and protect humane work in an increasingly automated and disconnected world.

Genres: Nonfiction, Sociology, Technology, Business, Audiobook, Leadership, Politics

384 pages, Hardcover

Published June 4, 2024

Reviews

[Kevin Postlewaite · Rating 2 out of 5]

Caveat: I quit about 1/4 of the way through.

There are many areas in which I am in agreement with the author:
-Personal connection/"seeing/being seen" is very psychologically important and can have a big impact on the outcomes of personal interaction
-Many interactions with workers have personal connection bundled in to the service interaction
-Many systems put in place by companies to regiment how employees do tasks can reduce the personal connection in a service interaction
-Most computer services these days do not support personal connection
-Reducing the amount of personal connection in a service interaction tends to disproportionately negatively impact those of low/lower socioeconomic status

Where I disagree with the author:
-that firms should be pushed to vaguely "be nice" and not specify how their employees do their work
-that society is better off with personal connection haphazardly and expensively interspersed through the economy, even those of low/lower socioeconomic status
-that computers and machines will never be able to form personal connections with people [on this, there is already an academic study where people rated LLM provided medical messages as more empathetic than those provided by people]

This is one of those books that identifies "what is good" but has terrible understanding with how this good can be realized in practice.

[Karen Adkins · Rating 5 out of 5]

Pugh's a sociologist, and she makes a compelling case for what she calls "connective labor," work that depends on relationality and mutual recognition. Education and health care are two of her primary sample industries, but her analysis extends to a whole host of other jobs. Because I'm a professor and know a lot of health care professionals, I find this idea pretty persuasive. The critical part of her argument is urgent. She claims that late capitalism's enthusiasm for efficiency in work leads to exactly the wrong sorts of incentives (measuring, scripting, use of technology) applied as simply as possibly, that often serves to undermine exactly the sorts of relational labor that are most urgent (think of the doctors limited to 15-minute patient appointments, or the ever-increasing class sizes in education that cut across substantive engagement with students as full people). She warns us of possible futures where relational labor is available only to those with the ability to pay (see: concierge care in medicine, high-touch private schools), with most people getting rude facsimiles (chatbots "teaching" or providing medical advice), and ends with a clear exhortation of how advocacy for a different future could work.

[Henry Reed · Rating 2 out of 5]

The author has something important to say but it's not a book's worth of information and the reader must endure a lot of filler to get to the substance. The book seems unsure whether its readership will be the book club crowd or a more academic audience and ends up aiming in the middle, probably to the benefit of neither. I definitely gained some value of insight and perspective from the longer book and don't begrudge the hours spent reading it, but I suspect I'd love the TED Talk version.

[Peter Tillman]

From Nature's Five Best list, https://www.nature.com/articles/d4158...
"A healthy society requires more connective labour, not more automation, [sociologist Pugh] argues in her engaging study, which observes and interviews physicians, teachers, chaplains, hairdressers and more."

[Clark · Rating 4 out of 5]

Very good read. It has an ethereal complexity that makes it hard to think through all the practical applications for one's life, but at the same time it has incredible vignettes and is thoroughly thought provoking.

[Lieke vd R. · Rating 4 out of 5]

Een heel gevecht. Academisch Engels is net te hoog gegrepen voor mij denk ik. Desondanks wel enorm interessant, veel waardevolle tips uit het werkveld kunnen halen en mij tot nadenken aangezet over wat het betekend om als baan “mensen te zien”. Een aanrader voor wie ‘in de mensen’ werkt.

[Mary · Rating 4 out of 5]

Research into how the drive for efficiency is eliminating connective (empathic/compassionate) work. Soon only the rich will have access to human labor while the rest of us make do with apps and tele health. In the end she says we can fight this by demanding our medical folks have the time to share their care with us. So use the fill check out line folks! Stop using the self-checkout counter!

[Amy ☁️ (tinycl0ud) · Rating 4 out of 5]

Recommended by Prof Teo You Yenn

This book attempts to put into words that invisible spark of connection that is very important in achieving any kind of outcome involving human stakeholders, but which is also difficult to mass-produce, though not for lack of trying. The term used here is 'connective labour,' which is evocative of emotional labour, and I know the word 'labour' has certain connotations (is labour not something that gets extracted out of workers who cannot fight back?) but I liked that the way it is used here suggests that, if done right, the work itself can be meaningful and fulfilling for both the worker and the client. On the whole I felt that there were many useful concepts introduced here that can help the reader make sense of what is happening right now. If ever you felt a kind of ambivalence about people using chatbots for self-therapy, or calls for teachers to provide more 'care' for their students/healthcare providers to be more 'caring' towards their patients, or self-checkout kiosks at the supermarket, then this book would be helpful for you.

Some notes:
(1) I really appreciated that this book did not shy away from the fact that interactions between humans, no matter how well-intentioned, unavoidably produce harm, sometimes simply by accident (a "misrecognition" or misreading as "we are none of us perfectly legible to the other"), but that is not a good justification for avoiding connection. The "rupture" can function as a learning moment to deepen connection. Another possible outcome is the triggering of shame, which is why they may prefer to vent to a chatbot instead of a human therapist who they assume may be secretly judging them. I can sympathise with this view, but at the same time I'm wondering if it might be a good thing to feel shame in some situations. Or maybe that's just me being Chinese and thinking that having a "spectre of shame" is evidence of a moral compass, as opposed to being shameless/无耻.

(2) Brought up as well is the common assumption that this kind of labour is "associated with femininity, presumed to be part of women's nature" but I think (and it is intimated here as well) being able to care well for another, like listen well and connect with them on a personal level, is a skill that can be learnt by any person of any gender.

(3) The idea that automation can bridge the gap between unequal classes is clearly fiction. It would be awfully convenient if the masses could placidly accept "being seen by machines" following pre-programmed scripts because then they won't make their discontentment their governments' problem. This book suggests that the gap is possibly widened even further as "being able to have a human attend to your needs has become a luxury good" reserved for the wealthy who can afford to book out a service provider's entire day, but that has also led to the creation of a "servant class." A line on this here: "If we continue to prioritise efficiency over relationship, we create a new kind of haves and have-nots: those divided by access to other people's attention. Without addressing the conditions of connective labour, we foster the stratification of human contact."

(4) On a larger scale, e.g. "organisations with corporate social architecture" that has more resources, workers are still being worked to their maximum rather than having more time because the clients inhabiting such upscaled spaces also have upscaled demands. The culture there also "involves a consumerist ethos" so service providers have to collect data to improve metrics like customer satisfaction. It's so counter-productive.

(5) For the service provider (teachers, nurses, therapists, etc), their working conditions improve if they take an "artisanal" approach to their careers and cater to fewer but higher-paying clients who can afford them, but often that goes against the reason they chose their careers in the first place, which was out of a desire to serve the greater good. Their skills and expertise no longer benefit those who actually need it most (the poor and disadvantaged) but, sadly, to the highest bidder. The alternative is to return to understaffed institutions where each worker is being squeezed of every bit of value they can produce until they die inside or they burnout and quit.

(6) I found interesting the case studies when people tried to do a human-assisted technology or technology-assisted human approach, basically when an organisation tries to see what they can do to make up for the other's shortcomings. I agree that all the desirable skill sets we want from a single teacher is quite impossible to be "traditionally found in the same person" so if you want a great coach then maybe he isn't the best math tutor. But honestly I don't think letting kids do all their learning independently from a tablet is great, and my innate response is fear and distrust, but it does free up the teacher. And yet, the teacher still ends up being a data-logger. Is there any real value to collecting so much data on each child? I wonder if maybe the goalposts themselves are the issues.

(7) The section towards the end about SEL programmes had this line that hit hard: "these SEL programmes assign the children individual responsibility for feelings that arise out of socially structured inequalities" > why does this sound just like the Peer Support Network thing? These accounts of teachers having to check-in with the students' emotional states but then being forced to quickly "subordinate them to meet the dictates of [their] employer" is so true. Surely it's more harmful to make students share their emotional state but do nothing about it and continue on with the lesson because class time is precious? Why is the lesson always more important than the child's wellbeing?

(8) On a more personal note--
As I was reading, I recalled when a student told me that they liked me because [Point] I was the only adult who seemed happy to see them. [Evidence] I would wave back at them, say hi, and say "see you next week" in an enthusiastic tone. [Analysis] If they liked me because they were under the impression that I liked them, possibly it means that if they felt as if I hated being in their presence, I would then be hated in turn. If I did not smile, laugh at their jokes, play along, dole out forgiveness, etc, the assumption then would be that I was not happy to see them, so I would not have 'earned' their favour. The simple question is: was I pretending? No. I don't have the kind of energy to pretend for weeks on end like some kind of professional actress, and also they were genuinely very lovable so I had no issues. I had no issues all the way because it was cooperative and collaborative. I felt that we understood each other perfectly. But the real question is: if I did not like them, was it my duty or obligation to pretend that I did? Putting aside the obvious positive academic outcomes (improving grades, decreasing behavioural issues, increasing self-esteem), does it count as deception or not? [No Link?]

(9) In many of the case studies shown in this book, the people who chose to provide connective labour often did so willingly, because they were already naturally inclined to and entered the profession because it was a calling and helping people felt like their life's mission. One person even said that he filtered out teachers who did not look like they loved kids in the hiring process. There were also people, not interviewed but mentioned in the background, who did not. They saw clients (customers/students/patients/etc) as a "rotisserie chicken" to be turned, machines to be fixed asap, walking dollar signs, etc, and themselves as mere workers to deliver a service or product as minimally as possible. I'm still struggling to figure out where I stand on this, because I have been on both ends, caring a lot and then taking on individual suffering, but also caring very little but feeling like an assembly line robot, and I'm not sure going down the 'artisanal' route would be a solution for me personally because I would feel like nothing more than a glorified PA or private tutor, still a servant. There has got to be a fourth way, or a fifth way.

[Meghan · Rating 3 out of 5]

Overall I appreciated the idea of this book, I just might have preferred it as an article or shorter read.

[Jennifer Sertl · Rating 5 out of 5]

The Last Human Job by Allison Pugh reviewed by Jennifer Sertl
Allison Pugh’s The Last Human Job is one of the most important books I have read on the future of work. At a time when conversations about AI tilt toward efficiency, scale, and automation, Pugh brings the discourse back to what truly shapes our experience of work: our relationships. Her writing is grounded, deeply researched, and refreshingly human. It reminds us that no matter how advanced our tools become, connection is still the essence of meaningful labor.
A Book That Centers Connection
Pugh, a sociologist and former president of the American Sociological Association, has spent decades studying how people maintain dignity and meaning in environments shaped by economic pressure and technological change. What she illuminates with remarkable clarity is that there is an entire category of work—teaching, nursing, counseling, caregiving, coaching—that depends on the kind of attention machines cannot replicate.
She calls this “connective labor,” and the stories she shares demonstrate that it requires judgment, attunement, and a type of embodied presence that cannot be standardized.
Human-Centered Leadership
In our conversation, Pugh spoke about leaders who understand the stakes of designing cultures that support connection. One example was Burt Juster, a school principal who hired teachers based on their ability to build relationships first, trusting that curriculum could be taught later. For him, relational skill was not optional—it was the foundation for everything else. Under his leadership, practices like restorative circles became part of the school’s daily rhythm, demonstrating how culture is built through rituals that honor presence and listening.
These examples are reminders that empathy at work is not a sentiment but a skill. It requires leaders to design systems where people can slow down enough to actually see one another.
Technology That Serves, Not Substitutes
One of Pugh’s strongest arguments is that technology is never neutral. It can reinforce bias as easily as it can reduce workload. It can support connection—or sever it. She warns against the belief that more measurement always produces better outcomes. In some fields, such as healthcare, the push for quantification has eroded the very interactions that lead to healing.
I found her emphasis on trust especially powerful. In an era of automation, trust becomes a strategic asset. When digital systems are integrated thoughtfully, they can remove friction and support human judgment. When deployed carelessly, they reduce our interactions to transactions.
Risks We Cannot Ignore
Pugh outlines three emerging patterns that should concern every leader:
The triage effect: Machines handle the easy cases, leaving humans only the most complex and emotionally demanding work.
Unequal access to human attention: Those with resources receive real human care; others are routed to automated systems.
Divided labor: Technology takes on cognitive tasks while humans are left with emotional labor, widening class and workplace divides.
What makes the book hopeful is that Pugh does not resign us to these futures. She shows examples—across schools, clinics, and community organizations—where investment in relationships leads to better outcomes, stronger cultures, and even cost savings.
A Call to Reclaim Our Humanity
The Last Human Job is ultimately an invitation. It asks us to design workplaces where empathy, mercy, and dignity are not afterthoughts but essential infrastructure. It asks leaders to see connection as a competency. And it asks all of us to remember that the most meaningful parts of our work cannot be automated.
In an era that often feels rushed and disembodied, Pugh’s message is a grounding one:
Technology can assist us, but only humans can care.
For anyone thinking about the future of work—leaders, educators, healthcare professionals, technologists, or anyone navigating change—this book provides both an anchor and a path forward. It left me more hopeful, more attentive, and more committed to preserving what makes our work feel human.

[Willi V · Rating 4 out of 5]

The Last Human Job: The Work of Connecting in a Disconnected World. I went into this book expecting it to be about friendship and was surprised to find a sociological account of the condition of what the author calls "connective labor." Connective labor is work that is largely based on building relationships with one's clientele: e.g., psychotherapy, medicine, nursing, teaching, sales, hairdressing. The book is based on meticulous field observation and interviews and paints a thorough picture by exploring the following topics: why connective labor is important, how it works, how it is being automated and systematized, under which conditions it is hampered and under which it flourishes.

A major strength of the book is how it works largely unideologically. It takes into account the socioeconomic reality of the current age USA and draws conclusions based on how it affects connective labor without passing moral judgment or being overly polemic.

However, the chapter on how AI is being used to do connective labor leads to a short-sighted conclusion in favor of the necessity of humans in this field.

You may be able to get something out of this book if you are: 1) a connective laborer interested in how certain working conditions affect your work and your life, 2) interested in the social architecture of health services, or 3) interested in what techniques or takeaways you can get for your own work which, if you work with humans, may involve some inkling of connective labor as well. However, be prepared for a somewhat redundant and boring reading experience.

[B · Rating 3 out of 5]

2.5

看完還是對「連結性勞動」這個概念有許多困惑。作者將連結性勞動定義為「透過與他人鍛造情感的理解，從而創造我們認為重要的那種結果的工作」，與此同時強調連結性勞動的關鍵在於「看見他人／被看見」。作者以「概念性民族誌」是試圖挖掘這個概念，因此訪談主要對象為治療師、教師和主要照顧醫師。

我自己對於這個方法是否能幫助我們理解連結性勞動有許多懷疑。我對概念性民族誌並不熟悉，但若概念性民族誌是「概念走在研究前端，前往他所引領的各種場域」，我會覺得作者劃定了主要三個職業別某種程度上是已經對概念有先驗的成見，卻也沒有試圖透過經驗材料來修正自己的成見。我在閱讀這本書的過程，會一直困惑連結性勞動是只侷限在特定領域嗎？但依據作者在導論提到的，連結性勞動是一層在許多勞動下的額外勞動，那麼私以為想要實踐概念性民族誌的方式，應該是去訪談百工百業，提煉出共同的，以「肯認」「關係」為主軸的勞動。

我在讀這本書的另外一個困惑是，連結性勞動和情緒勞動究竟有什麼差異？我嘗試解讀出的是或許情緒勞動並不一定要存在「肯認」的要素，但我認為在定義中放入「重要的那種結果」或是「被看見」等涉及主觀感受的要素並不是一個能幫助理解了解概念的好方法，且若「被看見」真的這麼重要，我認為另一條理解這個概念的路徑，應該像是 Hochschild 一樣專注於其中一種職業別，去談說工作者和服務對象（總之，「被看見的人」）怎麼理解這個勞動。只劃定三個主要職業，卻又把他們混為一談，會讓人產生這個概念是指侷限於特定產業嗎？以及忽略各個職業別各自的內涵的困惑。

或許是本書的台灣譯本以「AI 浪潮下最難被取代也最受威脅的工作」行銷，讓我對這本書有錯誤期待。的確有觸及自動化浪潮，但並不是很多。本以為是回應當代社會現象的書，但所反省工作腳本化、追求效率帶來的負面後果等都早有許多勞動研究或是建制民族誌所討論過了。雖然宣稱是一本以當代數一數二重要的社會現象為背景的書，然而整體研究成果都讓人感覺不到新意。

[Rochelle · Rating 4 out of 5]

I first heard Allison Pugh on NPR's Hidden Brain talking with Shankar Vedantam about her research into connective work. It was her peaceful presence that came through the interview which propelled me to read this book. Hundreds of interview hours have gone into this book as she seeks to deeply inquire into the nature of connective work as more and more disciplines and professions, and especially the professions to which establishing human connections are fundamental, are being taken over by AI functions and are increasingly subjected to the algorithms of efficiency and dollar cost averaging and profit. I think anyone, anyone, who does any kind of work with people should read this.

[Laura · Rating 5 out of 5]

This is a profound work documenting the value of connective labor and what will be lost if we abandon it to AI. Pugh expresses what humans do for each other with greater precision than the simple “empathy” or “community” characteristic of much writing on this topic. Empathy and community are good ideas, but they can be vague—Pugh’s use of apt examples and in-depth analysis makes a more powerful case than I typically see.

[Emilia Greendevald · Rating 5 out of 5]

I’ve noticed many recruiters appreciate resumes that are easy to scan. That’s exactly what a resume builder https://www.gotresumebuilder.com/ helps you achieve - clarity and readability. By using structured templates, your key information stands out immediately. This is crucial because hiring managers often spend less than a minute on each application. With a tool like this, you make every second count.

[Vjera · Rating 4 out of 5]

3.5 stars. The concept of the importance of connective labor was valuable and interesting. The examples, particularly from the medical facility, were interesting and engaging. It got me thinking and talking about connective labor. But some sections were very dense, it felt almost textbook heavy in places. And the way quotes were written drove me crazy, with all the starts, stops, and switches that we have in live speech but tend to get cleaned up in a book like this.

[Divyesh Sheth · Rating 4 out of 5]

Book raises profound questions around human connection and how the work that requires human touch have evolved and attempt to standardize them and pursuit of profits and efficiency have changed these jobs and how we can preserve the human touch, dignity and social cohesion in the age of AI and automation.

[Eliza · Rating 4 out of 5]

This is a seminal book which all leaders should read to understand how important humans are to interpersonal work. Pugh sheds new light on this subject and uncovers exactly why humans, and not machines, must carry on doing interpersonal work.

[Olyvia · Rating 1 out of 5]

Read this for university.

This book could have been 303 pages shorter. If you like books or any form of media for that matter that repeats the central topic and its applications like you have amnesia, this one’s for you.


0 stars if I could, if that wasn’t already clear.

[Thomas Estella · Rating 2 out of 5]

Extremely thought provoking subject. If the conclusion was formatted as an op ed that was expanded just a little more that would be the perfect length. A topic worth discussing, but the drawn out chapters almost made me shelf the book.

[Kate Hornstein · Rating 3 out of 5]

This book would have benefited from some subheads within each chapter.

[Lucas Henderson · Rating 5 out of 5]

My first book of the year but my must read of the year.

Pugh really hammers down that what matters is the human, and the ability of seeing the other and being seen.

[Roy Kenagy]

Prince UP Announcement

[Anna]

A listened to the 1st half with me.

[Suzy G · Rating 5 out of 5]

Extremely thought provoking as someone who does, what Allison Pugh calls, "connective labor" everyday. The stories interwoven in the book are insightful and make the text unique.

[Finn · Rating 5 out of 5]

Connective labor makes the world go round !

[Kira Dawson · Rating 4 out of 5]

You wouldn't guess it from the title, but I don't think this sets out to be a big anti-ai/automation book. It gets there though. A good book for the hesitant.

[kiran · Rating 4 out of 5]

very good. now i am very worriedly rethinking my career plans in computer science though

[Matt Singer · Rating 4 out of 5]

Really timely read, almost wish she had extrapolated further I think this is applicable to a lot of workplaces.

[Marc Schultz · Rating 4 out of 5]

Long but wall thought out