Suicidal Empathy: Dying To Be Kind, Gad Saad
A final review on empathy, forbidden knowledge, criminality, ideology, and other cognitive bugs of the West.
The core thesis is powerful in its simplicity: empathy is a virtue, but like all virtues it can become dysfunctional when applied in the wrong way, toward the wrong target, with the wrong intensity.
There are books that do not merely defend a thesis. They force you to recalibrate the measuring instrument with which you observe reality. Suicidal Empathy: Dying To Be Kind by Gad Saad belongs to this category: not an accommodating book, not a gentle book, but a necessary one.
…many objections seem to confirm exactly the diagnosis of the book: when a datum contradicts the narrative, the datum is declared morally contaminated. When criminal behavior cannot be reduced to structural victimhood, the conversation changes. When reality is not empathetic, reality is put on trial.
For those who know me, this sociological detour away from the rails of engineering and information theory may seem strange.
In reality, for me the passage is much less exotic than it appears. For years I have worked with data, models, retrieval, predictive systems, and information architectures. A large part of my work consists in distinguishing signal from noise. And the problem Saad addresses is exactly this: what happens when a society loses the ability to distinguish an authentic moral signal from emotional noise amplified to the point of distortion?
1. Empathy as a virtue out of calibration
It is not enough to say that thought can be colonized by ideological dogmas. Emotions can be hacked too.
The core thesis is powerful in its simplicity: empathy is a virtue, but like all virtues it can become dysfunctional when applied in the wrong way, toward the wrong target, with the wrong intensity.
This is not a denial of empathy. It is, rather, a critique of empathy without a cost function. And anyone who works with models knows very well that every optimization function without constraints produces mathematical monsters. Elegant ones, perhaps. But monsters.
The book presents itself as the almost natural continuation of The Parasitic Mind (2020). There Saad analyzed the parasitic ideas that infect cognition. Here he shifts attention to the emotional module. It is not enough to say that thought can be colonized by ideological dogmas. Emotions can be hacked too.
Empathy, in this perspective, becomes an exploit: a noble vulnerability used by cultural, political, and media systems that reward moral posture more than real world outcomes.
2. Forbidden Knowledge: when truth becomes impolite
One of the strongest parts of the book, and the one I would frame in the most academic terms, is Chapter 2, Forbidden Knowledge. Saad starts from Roger Shattuck and his taxonomy of forbidden knowledge: knowledge that is inaccessible, prohibited by authority, or simply dangerous because it is unwelcome. The most interesting point is not the taxonomy itself, but the mechanism Saad attaches to it: in many contemporary cases, the refusal to investigate uncomfortable truths is justified as a form of empathy.
This is a decisive move in the argument. The issue is no longer only that certain ideas are wrong, offensive, or poorly expressed. The issue is that some questions should not even be asked, because asking them might harm someone, destabilize a protected narrative, or produce socially undesirable consequences. Here empathy mutates into an epistemological censor.
Saad uses the example of whether one should tell a friend an unpleasant truth, and places the question between deontological ethics and consequentialism. Do we tell the truth because truth has intrinsic value? Or do we conceal it because its disclosure may hurt someone? Translated into academia, the problem becomes much larger. Research areas can be discouraged not because they are methodologically weak, but because their possible conclusions are morally inconvenient.
This is where the book intersects strongly with my own professional instincts. A society that refuses to query the database because the output may be unpleasant is not compassionate. It is epistemically fragile. It is building dashboards in which the most important indicators are hidden because they do not comply with the desired story.
3.1 “Blank Slate Felons”: the criminal as tabula rasa
Chapter 4 remains, in my view, one of the central nodes of the book. “Blank Slate Felons”: criminals as blank slates.
Saad attacks head on the idea that the criminal is always and only the passive product of external conditions: society, poverty, discrimination, trauma, context. All true, as long as these remain part of the model. False, when they become the entire model.
Here my instinct as a data person wakes up immediately. A model that eliminates an entire class of variables because they are politically inconvenient is not compassionate. It is bad modelling. If we erase individual predispositions, behavioral traits, personal agency, biological differences, personality structure, recidivism, and predictive signals, we are doing something very precise: we are overfitting ideology onto the data.
Saad summarizes this distortion brutally: through the lens of suicidal empathy, criminals are reframed not as moral agents, but as originally pure creatures corrupted by the social order. It sounds like a caricature only until we turn on the news, read certain commentaries on court cases, or listen to public debates about crime.
3.2 When the victim becomes a narrative detail
One of the most effective examples in the chapter is unexpectedly pop. Saad starts from an episode of Cheers, “Homicidal Ham”, in which Diane interprets the robbery she has suffered as a kind of cry for help from the aggressor. It is a comic scene, but it works because it exaggerates a real reflex: the temptation to search immediately for the hidden wound of the perpetrator before looking at the obvious wound of the victim.
The book then moves to real cases, much less comic. Saad recalls the murder of Dartmouth professors Half and Susanne Zantop, and from there reflects on parole, recidivist behavior, rehabilitation, and systemic indulgence. The point is not that every criminal is beyond redemption. That would be the opposite stupidity. The point is to ask what happens when a civilization begins to treat the victim as an emotional inconvenience in the redemption arc of the aggressor.
At that point we are no longer in the domain of justice. We are in badly funded experimental theater.
3.3 The problem is not rehabilitation. It is the causal model
Rehabilitation makes sense. Proportional punishment makes sense. Distinguishing between types of crime, contexts, intentionality, and risk of reoffending makes sense. But all of this requires something that the ideology of the blank slate cannot tolerate: a realistic causal model.
The theme of the tabula rasa is central. And here Saad connects, at least for me, with a tradition that has long interested me from different angles: Kahneman, with his biases and cognitive shortcuts; Gilbert, with our systematic errors in forecasting future emotional states; Anderson, with computational models of cognition; Marr, with the idea that a phenomenon must be understood at different levels (computational, algorithmic, and implementational).
Applied to human behavior, this means something very simple: you cannot explain crime by looking only at the social layer. You must also look at the underlying architecture.
This is what much contemporary ideology cannot tolerate: the existence of constraints. It prefers an infinitely rewritable human being, morally neutral, manipulable through policy, language, and reeducation. But the human being is not a JSON file to be fixed with a semantic patch. It is an evolutionary organism, with instincts, differences, limits, trade offs, incentives, and strategies.
4. Predictable criticisms, and why they are not enough
Naturally, Saad is accused, and will continue to be accused, of oversimplifying, of being right wing, of using evolutionary psychology as a cultural weapon, of confusing empathy with Western guilt.
Some criticisms must be heard. The risk of turning a strong intuition into a universal key always exists. Reverse scientism (the kind that claims to explain everything with one variable) remains scientism, even when it stands on the opposite side of political correctness.
But many objections seem to confirm exactly the diagnosis of the book: when a datum contradicts the narrative, the datum is declared morally contaminated. When criminal behavior cannot be reduced to structural victimhood, the conversation changes. When reality is not empathetic, reality is put on trial.
This does not mean that every page of Saad’s book is equally persuasive. Some passages have the rhythm of polemic more than the rhythm of analysis. Some examples are deliberately provocative. Some readers will find the tone too sharp, and not always without reason. But the basic diagnostic question remains intact: can a society survive if its moral sensors are calibrated to reward the display of compassion more than the consequences of compassion?
5. Islamism, Hamas, Israel: distinguishing without anesthetizing
On immigration, integration, criminality, antisemitism, political Islamism, and Hamas, my point is clear: we can no longer afford analyses based on performative emotion.
Criticizing political Islamism, its supremacist components, and its incompatibility with certain liberal structures does not mean dehumanizing Muslims as individuals. It means refusing the blackmail according to which every cultural analysis must be anesthetized in advance so as not to disturb the progressive catechism.
Likewise, my pro Israeli position does not prevent me from criticizing Netanyahu when I believe his choices fuel cycles of violence rather than produce long term strategic security. In my view, and in the view of many voices within Israel’s own security establishment, this is not a marginal concern. But here the point returns: criticizing a political leadership is different from denying Israel’s right to exist, and it is different from romanticizing theocratic and terrorist movements as if they were NGOs with a communication problem.
Get Marco Goldin’s stories in your inbox
Join Medium for free to get updates from this writer.
This is also where one of the most delicate passages in Saad’s book deserves to be handled with care. Saad recalls his own family history: Lebanese Jews forced to leave Lebanon, parents kidnapped by Fatah, grandparents who had fled Syria because of antisemitism, Egyptian Jewish relatives who left Alexandria, and Armenian relatives connected to the history of genocide and displacement. The point is not autobiographical exhibitionism. The point is that the contemporary hierarchy of victimhood often does not know what to do with Jews, especially Middle Eastern Jews, because they do not fit neatly into the simplified maps of privilege currently used in many academic and corporate settings.
Saad’s phrase “The Orwellian chutzpah is simply astounding” is harsh, but the issue it names is real. A ritual language of safety and inclusion can coexist with a striking indifference toward Jewish vulnerability. This is the paradox he identifies: institutions signal that they care, that they are empathetic, that they want a “safe space”, yet that safety may disappear when the subject is Jewish students, Jewish professors, or Israelis after October 7.
This must be discussed with gloves on, as we say in Italian, because public opinion in many spaces has become extremely polarized on Israel and Palestine. But gloves do not mean anesthesia. They mean precision. Hamas is not Palestine. Palestinians are not Hamas. Criticism of Israeli government policy is legitimate. Antisemitism disguised as anti Zionism is not. The distinction is intellectually basic, and morally indispensable.
6. AI, computational psychology, and common sense
Having now finished Suicidal Empathy, I can say that the book does maintain its analytical force, although not always with the same level of discipline. Its strength is conceptual clarity. Its weakness is that the rhetorical temperature sometimes rises faster than the evidentiary apparatus. This is not fatal to the book, but it matters.
For me, the book is interesting because it forces us to reason in terms of systems. A social system should not be evaluated by declared intentions, but by outputs. If a policy produces more victims, more insecurity, more crimes committed by repeat offenders, and more disintegration, it does not become good because the person proposing it had moist eyes and the right tone of voice.
This applies to information systems, to AI, to justice, and to politics. Intention is metadata. Result is data.
A model that optimizes for moral exhibition will produce moral exhibition. A model that optimizes for safety, truth, proportionality, and institutional resilience has at least a chance of producing a livable society.
— -
(If you read this far, thank you from the bottom of my heart)
many objections seem to confirm exactly the diagnosis of the book: when a datum contradicts the narrative, the datum is declared morally contaminated. When criminal behavior cannot be reduced to structural victimhood, the conversation changes. When reality is not empathetic, reality is put on trial.
Reading Saad without switching off the brain
As I said above, the book should ideally be read after The Parasitic Mind, because the two texts speak to each other: first the mind infected by ideas, then the emotion infected by a virtue out of calibration. It should be discussed not as a party manifesto, but as a systemic warning.
At the same time, reading Saad well means not outsourcing one’s judgment to Saad. This is important. A reader should not replace one dogma with another. The question is not whether empathy is good or bad. Of course empathy is good. The question is: toward whom, when, within what limits, and at what cost for those who have no voice inside the model?
In other words: fewer tears as user interface, more reality as cognitive engine.
A final critical note: a book born from a viral intuition
There is one final point that deserves an impartial note. In the acknowledgments, Saad explains that on March 19, 2024 he posted some reflections on X about suicidal empathy and the death spiral into which he believed it was pushing the West. Shortly afterward, Eric Nelson, executive editor at HarperCollins and publisher of Broadside Books, sent him the link to that post with the subject line “Here’s your book idea.”
The post is here: https://x.com/GadSaad/status/1770076895935152307
This origin story is fascinating, but it also raises a methodological issue. A concept born in the compressed, viral, polemical environment of X can become a powerful book idea. But virality is not validation. It is distribution. It tells us that a phrase touches a nerve, not that the full causal architecture behind it has been demonstrated.
This is where my final judgment becomes both appreciative and cautious. Suicidal Empathy is valuable because it names a real phenomenon: the miscalibration of empathy, the moral inversion between victims and aggressors, the tendency of institutions to confuse compassion with denial of reality. But the book is strongest when treated as a diagnostic lens, not as a closed theory of everything.
It is a warning signal, not the whole dashboard.
Mini bibliography for further reading
I strongly recommend reading Saad’s notes and following the bibliography behind the book. These are some of the works that help frame the discussion beyond the polemical surface:
Irene Blanken, Niels Van De Ven, and Marcel Zeelenberg (2015). “A meta analytic review of moral licensing.” Personality and Social Psychology Bulletin, 41(4), 540 to 558.
Rob Henderson (2024). Troubled: A Memoir of Foster Care, Family, and Social Class. New York: Gallery Books.
Maja Kutlaca and Helena R. M. Radke (2023). “Towards an understanding of performative allyship: Definition, antecedents and consequences.” Social and Personality Psychology Compass, 17(2), e12724.
Carolyn Pedwell (2012). “Economies of empathy: Obama, neoliberalism, and social justice.” Environment and Planning D: Society and Space, 30(2), 280 to 297.
Alan Levinovitz and Awais Aftab (2025). “The Rumpelstiltskin effect: Therapeutic repercussions of clinical diagnosis.” BJPsych Bulletin. doi: 10.1192/bjb.2025.10137.
Nicholas A. Christakis and James H. Fowler (2013). “Social contagion theory: Examining dynamic social networks and human behavior.” Statistics in Medicine, 32(4), 556 to 577.
Nicholas A. Christakis and James H. Fowler (2007). “The spread of obesity in a large social network over 32 years.” New England Journal of Medicine, 357(4), 370 to 379.
James H. Fowler and Nicholas A. Christakis (2008). “Dynamic spread of happiness in a large social network: Longitudinal analysis over 20 years in the Framingham Heart Study.” British Medical Journal, 337. doi: 10.1136/bmj.a2338.
Leslie P. Boss (1997). “Epidemic hysteria: A review of the published literature.” Epidemiologic Reviews, 19(2), 233 to 243.
Robert E. Bartholomew and Simon Wessely (2002). “Protean nature of mass sociogenic illness: From possessed nuns to chemical and biological terrorism fears.” The British Journal of Psychiatry, 180(4), 300 to 306.
Jeremy Rifkin (2009). The Empathic Civilization: The Race to Global Consciousness in a World in Crisis. New York: Penguin Books.
Stephanie D. Preston and Frans B. M. de Waal (2002). “Empathy: Its ultimate and proximate bases.” Behavioral and Brain Sciences, 25(1), 1 to 20.
Frans B. M. de Waal (2009). The Age of Empathy: Nature’s Lessons for a Kinder Society. New York: Crown.
Jean Decety (2011). “The neuroevolution of empathy.” Annals of the New York Academy of Sciences, 1231(1), 35 to 45.
Paul Bloom (2016). Against Empathy: The Case for Rational Compassion. New York: Ecco.
Barbara Oakley, Ariel Knafo, Guruprasad Madhavan, and David Sloan Wilson, editors (2012). Pathological Altruism. Oxford: Oxford University Press.
Fritz Breithaupt (2019). The Dark Sides of Empathy. Ithaca: Cornell University Press.
These works are worth reading because they prevent the discussion from collapsing into slogans. Empathy is too important to be left either to sentimentalism or to cynicism.
Thanks!
Thank you to everyone who has read this review, argued with it, shared it, criticized it, or simply used it as a prompt to think more carefully. That is the only real purpose of a text like this: not to close the argument, but to make the cost of lazy thinking a little more visible.
And, of course thank you, Professor Saad, for your work.
