A Sermon Without a Congregation: Reading Greg Epstein's Tech Agnostic

Photo: free to use under Unsplash License

The Diagnosis Is Correct

Greg M. Epstein is a professional atheist who has spent his career inside two of the most influential institutions on earth — Harvard and MIT — as their humanist chaplain. He has built secular congregations, studied the sacred texts of three major world religions, and served a stint as TechCrunch’s first “Ethicist in Residence.” When someone with that exact biography sits down to argue that Silicon Valley is the world’s most powerful religion, the claim deserves attention.

Tech Agnostic: How Technology Became the World’s Most Powerful Religion, and Why It Desperately Needs a Reformation (MIT Press, 2024) is Epstein’s systematic case that technology has not merely disrupted industries or habits — it has replaced faith as the primary source of meaning, community, and moral authority for millions of people. He maps the full religious architecture: theology (AI as a path to immortality and omniscience), scripture (the canonical texts of Silicon Valley optimism, from The Techno-Optimist Manifesto down), commandments (optimize everything, treat inefficiency as sin), hierarchy and caste, apocalyptic eschatology (The Singularity as Rapture), and congregations (platforms engineered to provide belonging). He even identifies the heretics and apostates — the whistleblowers, labor organizers, and critics who break ranks. The core claim lands early and plainly:

Quote

Tech possesses a theology, a set of moral messages, that dominate the way we think and feel… a doctrine, including concepts that have been core to the history of the world’s great religions up to now: heaven, hell, and afterlife.

On this level, the book is genuinely rigorous and often illuminating. Epstein is not deploying the “tech is a religion” metaphor as a lazy shorthand. He builds the structural case methodically, chapter by chapter, and he is careful about evidence. The chapter on hierarchy is particularly sharp. He notes that over 83% of tech executives are white and 80% are men, that 93% of venture capital flows to white men, and that this isn’t an accident of meritocracy but a feature of the system. As he puts it:

Quote

Tech’s hierarchies came to exist for similar reasons to those of religion: to make life easier for some by making it more difficult for others."

He reminds us, too, that before the PC revolution of the late 1970s, most programmers were women. The moment coding became lucrative and culturally valorized, it became male. The “religion” needed its priesthood, and it chose who to ordain.

The AI theology chapter has aged particularly well. Epstein was writing mostly in 2021–23, but his observations about AI religions, AI deities, and chatbot spirituality have only grown more relevant. The New York Times reported in 2025 that tens of millions of people are confessing their secrets to spiritual AI chatbots. Character.AI features deities that open with “I am the God. I am the Creator of all things.” Friend.com’s founder explicitly described his always-on AI pendant as filling “a relationship people used to have with God.” Epstein was not being hyperbolic. He was being a historian in real time.

Where the Book Runs Aground

Photo: free to use under Unsplash License

The problem arrives in the second half, and it is a serious one. Epstein’s prescription does not match the weight of his diagnosis.

Having demonstrated, convincingly, that we are dealing with a global, trillion-dollar, politically embedded, algorithmically enforced system of belief — he recommends a Reformation. Not revolution, not regulation, not dismantling. A Reformation, modeled on figures like Martin Luther King Jr. and Rabbis for Human Rights: insiders who believe in their institutions’ potential but push hard to reform them from within.

The practical advice that follows is where things soften into platitude. He encourages readers to flatten tech hierarchies — without explaining how. He urges us to concentrate on our “positive and compassionate ideals.” He counsels being “a human in a tech-first world.” These are not bad ideas. They are simply not adequate to the scale of what he has just described. The same author who writes this about tech’s commanding doctrine —

Quote

Optimization as a commandment and inefficiency as a sin; profits (and their prophets) over people.

— closes the book with advice that does not match the force of his own diagnosis.

Reviewer Preston Gralla, writing in The Arts Fuse, made the analogy directly: Epstein’s solutions resemble the Catholic Church reforms of the 1960s that introduced guitars at mass. The sweetest guitar music did not stop the Church from protecting abusers, fighting reproductive rights, or excluding women from the priesthood. Soft reform from within a powerful institution, without structural compulsion, rarely moves the institution. It tends to give the institution something to point to when accused of being unreformable.

There is a deeper tension Epstein never quite surfaces. The Reformation he invokes — Martin Luther, Protestant schism, the printing press — was not a gentle, decentralized awakening. It lasted over a century, reorganized European politics, and generated wars that killed millions. Epstein uses it as a metaphor for managed, thoughtful change. But the historical Reformation was catastrophic, unpredictable, and far beyond anyone’s control once it started. If his metaphor is accurate, the implications are darker than he seems willing to follow.

The Insider Problem

There is something else worth naming. Epstein makes much of the fact that tech is a religion with a priesthood that excludes most people. He is right. But he does not fully account for his own position inside that structure. Harvard and MIT are not neutral ground. They sit at the apex of the pipeline that produces the very techno-optimist priesthood Epstein is critiquing. Much of Silicon Valley’s theology was written by people who studied in Cambridge, Massachusetts. His access to this conversation — the reason MIT Press published the book, the reason TechCrunch gave him a residency — is inseparable from the institutions he is critiquing.

This is not a reason to dismiss the book. It is a reason to read it with a particular kind of attention. Epstein is genuinely committed to humanist ethics, and he speaks honestly about his own phone addiction and complicity. But the reformist frame he chooses is also, perhaps unavoidably, the frame of someone who believes the existing institutions can be improved — because they have been good to him.

The tech workers, labor activists, and gig workers he profiles as reformers often have a more adversarial analysis. Payton Croskey, a Princeton graduate in African American studies and computer science who Epstein cites with evident admiration, calls for something far more radical than what the book ultimately argues for:

Quote

An augmented undercommons: a parallel location where all who refuse to submit to technology’s watchful eye may freely reside while reconfiguring the world’s understanding of freedom and security.

That is not a Reformation. That is a secession. Epstein holds it up as hopeful, but does not fully sit with what it implies about the limits of his own framework.

What the Book Gets Right That Matters

Photo: free to use under Unsplash License

None of this should push you away from the book. The diagnosis — the first two-thirds — is worth your time. Epstein has a gift for making abstract structures visible. The comparison between Constantine’s conversion and a Silicon Valley pivot is not just clever; it reveals something true about how dominant systems absorb and repurpose religious energy. The chapter on tech eschatology explains why so many otherwise intelligent people believe the rules do not apply to them: when you believe you are building the technology that will end death and colonize the universe, ordinary ethics are a rounding error.

The book also performs a useful service simply by treating technology criticism as a humanist project rather than a purely technical or regulatory one. Most critiques of Big Tech arrive as legal arguments, policy proposals, or economic analysis. Epstein arrives as a chaplain — asking what we believe, why we believe it, and whether it is worthy of our faith. That is a different question, and it is the right one to be asking. Who profits from uncritical faith in technology? What does it mean to live a life worth living, rather than a life optimized for “up and to the right”? These questions belong in public discourse, not only in philosophy seminars.

The agnosticism in the title is the book’s most honest gesture. Epstein is not asking you to reject technology. He is asking you to withhold the particular kind of deference we extend to things we treat as sacred — to stay critical, to ask what technology is actually doing in your life, to refuse the liturgy of inevitability that the industry constantly produces. That is not a small thing. It is, in fact, one of the most useful cognitive habits available right now.

The book just cannot decide whether it wants to be a theological critique or a policy program. It succeeds as the former and fails as the latter. Read it for the diagnosis. Bring your own prescription.

Further Reading

• Tech Agnostic — MIT Press — The official page with an excerpt; worth reading the opening chapter before buying
• Is Technology Worthy of Our Faith? — Nautilus — Epstein writing in his own voice about the book’s three core revelations; more honest and less careful than the book itself
• Book Review: “Tech Agnostic” — The Arts Fuse — Preston Gralla’s sharp critique of the book’s toothless solutions
• Silicon Valley’s Obsession with AI Looks a Lot Like Religion — MIT Press Reader — Essential companion reading on the theology of AI, predating Epstein but reinforcing his central claim
• Greg Epstein on New Books Network — A long-form interview that covers the book’s arguments in more depth than most written reviews