Max Edward Perry reviews Charlotte Blease’s Dr Bot: Why Doctors Can Fail Us — and How AI Could Save Lives (Yale University Press, 2025), reflecting on its arguments in the context of today’s debates about technology and healthcare.
This is my second review of Charlotte Blease’s Dr Bot: Why Doctors Can Fail Us — and How AI Could Save Lives. In my previous review (Perry, 2026), I focused on the book’s relationship with historical work in the sociology of health and Illness. Here, I would like to spend some time unpacking the book’s logic and reflect on its relationship to our current moment.
Dr Bot is an important text. The book represents a certain spirit of the age; the zeitgeist of contemporary healthcare and the promises of Artificial Intelligence. The book is split into four parts: accessing care, doctor-patient interaction, diagnosing, and empathy. It builds the case for mechanised thinking robots replacing human doctors. Blease does not give us much insight into what these robots might look and feel like; indeed, her grasp of the material properties of this mechanisation is pretty loose throughout, but she is clear that AI machines think better than humans. Blease claims that the majority of a doctor’s work —anything involving the production of clinical knowledge— should be turned into machine work.
Ultimately, these are claims about how we should best organise healthcare, and they rest upon profoundly unsociological approaches. Like a growing number of books that sit in the ‘popular science’ section of our bookstores, Dr Bot pursues a sociological question via the scientism of the evolutionary and behavioural sciences. The text does this by collaging together psy-studies and behavioural studies of medicine with surveys and speculative studies of medicine. For example, much is made in the text about the vast expansion of potentially clinically relevant data through social media, smartphones and new wearable sensors. This discussion rests largely upon a study published in the Australian & New Zealand Journal of Psychiatry titled ‘Objective smartphone data as a potential diagnostic marker of bipolar disorder’ (Faurholt-Jepsen et al. 2019). This study, even in its own terms, returns relatively uneven results regarding its hypotheses, and suggests no truly actionable correlation. Perhaps more concerningly, there is no critical engagement with the methods of the paper, the notion of ‘objective’ in this regard, and the implication that the use of such data to predict mental health episodes would require permanent digital surveillance for those patients deemed to be at risk. Studies like this are woven throughout the book, and the accompanying footnotes sometimes raise questions about the depth of engagement with these sources or the broader implications they suggest. While the book does not always pause for critical reflection on these varied findings, it has nonetheless resonated within digital health circles (Dhairyawan, 2026; Ferreira, 2025; Lehman, 2025) and has been featured and reviewed widely in the popular press (Blease, 2025b; Dworkin, 2025; Rennison, 2025; Warburton, 2025; Wen, 2025). This reception speaks concerningly about the current state of digital health discourse, as it does about the popular influence of the humanities and critical social sciences.
*
The text’s four sections each paint human doctors as defective computers. Evolution has ‘ripped ancestral scars on human behaviour’ (7) and thus it is time to disbar our scarred corporeal bodies from the work of producing knowledge, to give this work to robots. As healthcare’s self-professed ‘Gadfly’ Blease claims to be unafraid of controversial thought, but I cannot shake the sense that this is an affected controversy. Indeed, beyond a taste for chromatic minds, Blease still recognises a need to retain ‘humanity’ in the clinic, albeit delimited and cordoned off from the complex epistemological work required in the production of medical truths.
Whilst this argument, that we should remove the human-from-the-loop, is further than most in digital health circles are willing to go, Dr Bot is broadly in step with UK government policy. In the 2025 ten year plan for NHS England, the government commits to making every hospital ‘fully AI enabled’ and ‘make AI every nurse’s and doctor’s trusted assistant, saving them time and supporting them in decision making.’ (Dept. of Health and Social Care, 2025). It is clear that AI is envisioned as a tool which should take the bulk of the cognitive work in medicine. Across healthcare, AI is depicted as a way of (1) reducing error, and (2) expanding the amount of time clinicians have to do ‘care’, which is effectively cleaved free from the burden of knowledge work. As the ten year plan puts it: ‘Scale the use of technology like AI scribes to liberate staff from their current burden of bureaucracy and administration, freeing up time to care and to focus on the patient.’ (Dept. of Health and Social Care, 2025). In a recent parliamentary debate on the NHS Health Bill, Sonia Kumar, Labour MP for Dudley said:
Now we must embrace the global technological revolution; from artificial intelligence to robotics, we must harness it to improve patient-centred care. Used well, technology does not replace humanity in medicine, but restores it, giving clinicians more time to care. (United Kingdom, House of Commons 2026. Col. 925)
Jeremy Hunt’s 2022 book Zero shared this sentiment. Though preceding the release of ChatGPT3 to the public, Hunt argued that technological and organisational improvements in the NHS could reduce error to ‘Zero’ — a truly astounding claim to anyone who has given even a moment’s thought to how difficult it is to create hyper reliable complex systems. Hunt, much like Blease, argues that doctors need to be freed from the use of their brains; in so doing, technology will deliver us to a sickness free utopia. Even critical reviews of that book did not take issue with this central premise (Clarke, 2022). It is a noteworthy coincidence that, in making their arguments, Hunt and Blease both uncritically commend the oracular virtue of venture capitalist Vinod Khosla, with skin in the game. Digital health discourse is, of course, bound to the influence of capital.
Technology and AI specifically, is thus framed not just as a mechanism for improving the epistemic operations of medicine, but also as a means of cutting into the authorial power of doctors to define illness itself. Clinicians themselves are aware of the bind that they find themselves in. Reflecting on a recent survey of their membership, Dr Anne Kinderlerer, digital health clinical lead at the Royal College of Physicians said:
‘The emphasis on “human in the loop” in order to ensure AI safety ignores the risk for this to instead become “the fall guy in the loop”, where the physician is only acting as a liability sink for the AI decision maker.’ (Kinderlerer, 2026)
Clearly, whether a human is ‘in the loop’ is less important than the question ‘who or what has done the thinking?’ What is interesting to me about Dr Bot is that the book offers such a thorough attempt to justify a position broad healthcare policy holds, that robots should do the thinking, but that healthcare is, as yet, unwilling to reckon with. Dr Bot is the prophet of a coming crisis.
*
It is hard to overstate how important evolutionary psychology is to this argument. The particular and remarkable human capabilities, such as they are (i.e., “pattern recognition”), might effectively have provided temporary tools for constructing a medical bureaucracy; however, they are now a liability. These liabilities are to blame for all manner of problems, from overreliance on intuition to undue deference. The following passage is typical of Blease’s writing:
‘Although signals of prestige were generally reliable in the hunter-gatherer environment in which these biases were selected for, nowadays mismatches between our modern environment and our old psychology demonstrate just how easily our Stone Age minds can short-circuit.’ (86)
Throughout the book, evolutionary accounts of human and machine intelligence rest upon analogies and a fundamental fuzziness regarding the actual engineering processes involved in machines. That is, the unknowable ‘design’ of human intelligence is rendered as objective, empirically grounded evolutionary fact —in my view, all too hastily— and the well-documented and historically contingent design of large language models are never empirically pursued.
‘Like talking to the internet on steroids’ (29) …but in a good way. That is about as good a definition of AI as we get in Dr Bot. Throughout, a whole raft of divergent technologies are abstracted and eulogised with relatively little engagement with the technical realities or the wider technical or social scientific literature on what the machines do. AI, Blease says, must be implemented with ‘…careful consideration of many factors and new trade offs’ (8); however, these factors are never meaningfully discussed. Instead, the book proceeds with an unsupported and uncritical account of AI as artificial (nonhuman), intelligent (without definition) and exponential (without evidence).
Blease does make a commendable effort to raise issues of access, equality, and racism in existing healthcare practice. Still, it is a testament to the dangerous determinism of this style of thought that 160 pages into a book about AI in healthcare, the author wanders into a discussion about race essentialism and attempts to weave through it in the following astonishing passage:
‘It is highly doubtful that our ancestors met “racially” different people, and there is no evidence that natural selection adapted our mindware specifically to code other people’s “race” or “ethnicity”. (160)
Whilst I am pleased to learn that my mindware is not preprogrammed for racism, that the author has to rely on such clearly tenuous assumptions about what our ancestors did to stand up an argument about why a chatbot might be a better doctor than your actual doctor without accidentally justifying racism should give us some pause for thought.
*
Beyond the evolutionary, we find occasion to meet with some thinkers that might be more familiar to those of us in the tradition of Sociology and the Social Studies of Science. Thomas Kuhn, for example, gets an early mention, as does Marshall McLuhan. However, the engagement with both is glib. McLuhan’s famous ‘Art is anything you can get away with’ is taken at face value and derisively applied to doctors, as though McLuhan did not venerate artists, and thought ‘getting away with it’ was a problem rather than an act of sublime creation. Kuhn is simplified beyond utility. The Structure of Scientific Revolutions ( Kuhn [1962] 2012) is used to evidence an ‘inbuilt defensiveness among scientists’ (157). This defensiveness evolved in science, Blease says, to protect against ‘fads’, but is now working against innovation. Whilst not entirely wrong, the account is so crude that it renders Kuhn’s thought useless. Blease coopts Kuhn to say that resistance against technology is an ill-placed startle reflex that must be overcome, completely negating Kuhn’s more nuanced account of how ‘normal science’ works and why ‘fads’ might be bad, actually.
Other experts, such as psychologist K. Anders Ericsson—whose ideas were popularised by Malcom Gladwell—social psychologist of empathy Daniel Batson, and evolutionary psychologists Joey Cheng and Joseph Henrich, are taken more seriously. Together these thinkers constitute the intellectual scaffold of the book and the basis of Blease’s evolutionary account of human limitation. What is most dangerous about this is that it renders healthcare a game of computation, in which answers to ailments exist and must be found using the best tools of calculation. To put this differently, illness is out there, and the doctor (and specifically the doctor’s brain) becomes a sensor to be optimised. Humanities and social scientific studies of medicine have long challenged this view, but Blease’s book demonstrates how such critiques have faded from popular consciousness. For example, Blease uses Ericson’s popular science book Peak: Secrets from the New Science of Expertise (Ericson & Pool, 2016) to describe the ‘science of expertise’ and the importance of ‘purposeful practice’. However, to speak in Blease’s evolutionary terms, the brain’s plasticity, neural pathways, grey matter, and ancient programming are topped out; ‘human-mediated healthcare’ is doomed because, whilst expertise is possible in music, sports, or foreign languages, medicine has transcended the capacities of the human. This is articulated through a strange and entirely imaginary anecdote about the self-critical capacities of a ‘peak performing’ F.B.I translator (Italian to English) which underwrites the claim that even an amazing and imaginary doctor (which Blease calls ‘Dr Amazeballs’) will never be able to reach ‘peak performance’ (124-127). I am sure scientism used to be less insipid. Music, sports and foreign languages are either soft or simple, and thus fine for human expertise; Medicine is serious business. All of this is deployed as a politically progressive argument that allows social and environmental accounts of performance to lead the dance between nature and nurture. Indeed, throughout the book, Blease is concerned about equality, sexism, and systemic bias and discrimination in medicine.
However, despite Blease’s implied progressivism, Dr Bot is a reactionary text. A separation of ‘art’ and ‘science’ might be familiar to doctors; it is a heuristic taught in medical school, and its separation is rarely seen as a political act. However, here it is deployed to say that humans are too flawed to adjudicate on the science of health. Determinism is fascistic because it assumes there is a correct way to produce knowledge; this justifies centralisation and totalitarianism as long as the engineering is right. It is concerning that a book that exemplifies so much of contemporary healthcare thought and policy brings us here.
Dr Bot takes the decidedly technocratic position that rational systems can irradicate error by removing corporeal, error strewn humans. It draws lines around the activities that can be brought low for human consideration (music, language, sport) and which belong to some other domain of judgement (pathology, science, equality). Care is soft, heuristic, and irrational and can be left to crafty humans; medical knowledge is hard, cerebral, and determined and cannot be left to shoddy human wetware. To improve healthcare for all, we must push back against such notions. Knowing and caring are not separate operations; to know is always to know humanly. We must reckon with that; it cannot be displaced.
About the author
Dr Max Edward Perry is a sociologist and postdoctoral researcher at the University of Edinburgh. He works on the ERC and UKRI funded project “DARE”, investigating ‘data driven healthcare’. Max’s sociological work has hitherto examined medical records, healthcare bureaucracy, and databases. His PhD thesis, ‘A Sociology of Medical Records’, was completed at the University of Bristol in 2024.
References
Blease, Charlotte. 2025 Dr Bot: Why Doctors Can Fail Us — and How AI Could Save Lives. Yale University Press.
Blease, Charlotte. 2025b. The Big Idea: Why We Should Embrace AI Doctors. The Guardian, August 31, 2025. https://www.theguardian.com/books/2025/aug/31/the-big-idea-why-we-should-embrace-ai-doctors.
Clarke, Rachel. 2022. Zero by Jeremy Hunt Review – This Is Going to Hurt. The Guardian, May 22, 2022. https://www.theguardian.com/books/2022/may/22/zero-by-jeremy-hunt-review-this-is-going-to-hurt.
Department of Health & Social Care. 2025. Fit for the Future: 10 Year Health Plan for England. Department of Health & Social Care. https://assets.publishing.service.gov.uk/media/6888a0b1a11f859994409147/fit-for-the-future-10-year-health-plan-for-england.pdf.
Dhairyawan, Rageshri. 2026. An Artificial Dichotomy. The Lancet (British Edition) 407, no. 10528. https://doi.org/10.1016/S0140-6736(26)00208-4.
Dworkin, Ronald W. 2025. Dr AI Will See You Now: Human Doctors May Be Fallible, but They Have a Tool That AI Medicine Will Never Have: Judgement. New Statesman 153, no. 5831.
Faurholt-Jepsen, Maria, et al. 2019. Objective Smartphone Data as a Potential Diagnostic Marker of Bipolar Disorder. Australian and New Zealand Journal of Psychiatry 53, no. 2: 119–28. https://doi.org/10.1177/0004867418808900.
Ferreira, Becky. 2025. The Ascent of AI Therapists. MIT Technology Review, December 30, 2025. https://www.technologyreview.com/2025/12/30/1129392/book-reviews-ai-therapy-mental-health/.
Kinderlerer, Anne. 2026. AI Adoption Is about Leadership, Not Just Tech. NHS Alliance, May 6, 2026. https://thenhsalliance.org/resources/ai-adoption-is-about-leadership-not-just-tech.
Kuhn, Thomas. [1962] 2012. The Structure of Scientific Revolutions. 4th ed. Chicago: University of Chicago Press.
Lehman, Richard. 2025. Book Review: Dr Bot: Why Doctors Can Fail Us — and How AI Could Save Lives. BJGP Life, 2025. https://bjgplife.com/book-review-dr-bot-why-doctors-can-fail-us-and-how-ai-could-save-lives/.
Perry, Max Edward. 2026. Review of Dr Bot: Why Doctors Can Fail Us—and How AI Could Save Lives, by C. Blease. Sociology of Health & Illness 48, no. 2. https://doi.org/10.1111/1467-9566.70152.
Rennison, Nick. 2025. AI Is the Future of Medicine – Studies Show Almost Half of Doctors Can’t Cope with Their Workloads and Most of What They Learnt at Medical School Is Out of Date by the Time They Graduate. Daily Mail, September 5, 2025. https://www.dailymail.com/home/books/article-15065677/Dr-Bot-charlotte-blease-science-book-review.html.
United Kingdom. Parliament. House of Commons. 2026. Health Bill. Hansard Parliamentary Debates, vol. 786, June 1, 2026. https://hansard.parliament.uk/commons/2026-06-01/debates/A2EAF490-72AB-4F7E-998F-3108BD40F89A/HealthBill.
Warburton, Nigel. 2025. Everyday Philosophy: Dr Bot Will See You Now. The New World, October 1, 2025. https://www.thenewworld.co.uk/nigel-warburton-everyday-philosophy-dr-bot-will-see-you-now/.
Wen, Leana S. 2025. ‘Stethoscopes in the Sand’: Why I’m Rethinking AI’s Role in Medicine. Washington Post, October 7, 2025.
