Why Yuval Harari Is an Intellectual Bullshit

### On Yuval Harari's method - and on his lecture "Why advanced societies fall for mass delusion"

Before arguing with what Harari says, it is worth examining how he thinks and writes. Because he does have a method: stable, recognizable, repeated from book to book. Much of the strength, and the weakness, of his writing comes precisely from that method, rather than from any particular thesis.

## I. The Method: A Machine for Grand Synthesis

Harari's method is stilted in the exact sense of the word: it stands on a repeatable structure. The pattern is the same in *Sapiens*, *Homo Deus*, *Nexus*, and in his public talks:

1. Treat all of human history as a single field.
2. Find one master key: a single lens that explains everything (cooperation through fiction; later, information networks).
3. Present it in polished aphorisms, delivered with the confidence of established fact, almost without caveats or footnotes.
4. Add a counterintuitive twist ("in reality, X is Y") that functions as the engine of "depth."
5. End with a dramatic forecast about the future, pronounced with the authority of a historian.

This is effective: it becomes accessible, memorable, and easy to quote. His talent for synthesis and popularization is real, and denying that would be dishonest. But the structure has a critical flaw, and specialists have been pointing to it for a long time.

The main vulnerability is its unfalsifiable scale. His own doctoral supervisor at Oxford, Professor Steven Gunn, noted that in *Sapiens* Harari essentially slipped past fact-checking by asking questions so large that no one can point to exactly where he is wrong. This is his signature move: scale as a shield. When a thesis covers millennia and species all at once, it can be neither confirmed nor refuted. It is not a claim; it is an atmosphere.

The anthropologist Christopher Hallpike put it even more sharply in his review of *Sapiens*: he found no serious contribution to knowledge in the book - "where Harari's facts are broadly correct, they are not new; and where he tries to make a point of his own, he is often wrong, sometimes seriously." The list of scholarly critics is long: the neuroscientist Darshana Narayanan, the anthropologists Marlene Zuk and Mike Wilson, the philosopher Galen Strawson, and the science journalist Charles Mann. The common theme is the same: overgeneralization, the probable presented as settled, nuance sacrificed to the beauty of the arc.

And there is a universal rhetorical maneuver that makes the whole scheme almost invulnerable: "everything is stories/fictions." Applied this broadly, it becomes unfalsifiable and, more importantly, self-refuting: if everything is fiction, then so is his own grand narrative. More on that below, using a concrete example.

## II. A Close Reading: "Why advanced societies fall for mass delusion"

What he claims. The problem is not human nature but information systems: "people are generally good and wise, but give good people bad information and they will make bad decisions." The quality of information has not improved over thousands of years; modern societies are just as susceptible to mass delusion as Stone Age tribes. Large systems rest on an "alliance of mythology and bureaucracy"; "systems need a little truth and a lot of fiction." Information matters because it **connects** people, not because it reflects reality. Totalitarianism and democracy are different information networks (the second can correct mistakes, while the first, by default, cannot). And AI is an "alien intelligence" that invents stories and makes decisions on its own, and may strengthen totalitarianism.

What is genuinely strong here (because without this, the critique would be dishonest). "Fiction coordinates cooperation" and "mythology plus bureaucracy" are working sociology, even if they are not his inventions (Durkheim, Anderson's "imagined communities"). The frame of "democracy versus totalitarianism as different regimes for handling error" is a genuinely useful lens, even if it is hard to prove. The concern that AI scales propaganda is mainstream, not nonsense. This is not charlatanism.

**Where is the obvious demagoguery?**

1. A self-flattering false fork. "People are good and wise; information is to blame" is a comfortable fork that vindicates the audience and shifts everything onto an abstract "system." Nazism and Stalinism were not only "bad information"; they were also motivated reasoning, the will to power, tribal hatred, self-interest, and cowardice. And here is the irony: while warning that societies are susceptible to flattering stories, he himself offers a flattering story - "you are good and wise" - which means he is using exactly the mechanism whose dangers he is preaching about.

2. Unfalsifiability. "Good people plus bad information equals bad decisions," as he uses it, is almost a tautology: any failure can be retroactively written off as "bad information," while "good people" are simply postulated. It explains everything, which means it explains almost nothing. It is the same "too-large question" move as in the books.

3. An aphorism without a claim to precision. "A little truth and a lot of fiction" sounds crisp, but it is quantitatively empty: how do we measure that proportion? The formula performs explanatory work it has not earned.

4. Internal contradiction. "The quality of information has not improved in thousands of years" directly conflicts with his own claim that "democracies can self-correct." The scientific method, statistics, peer review, and a free press are precisely improvements in the quality of information. "We are just as susceptible as a Stone Age tribe" is a provocative simplification that erases exactly the institutions he praises one paragraph later.

5. A relativism that undermines his own project. If information matters not because of truth but because of connection, and democracy is "just another network," then on what grounds is it better? To defend self-correction as a good, he needs the value of truth - but his frame reduces truth to connective tissue. He is sawing off the branch he is sitting on.

6. "Alien intelligence" is branding in place of analysis. The label frightens, but it smuggles in contested assumptions: agency, autonomy, "alienness." The leap from "AI generates text" to "AI makes decisions by itself and weaves enslaving myths" is delivered with a confidence the data do not justify: the signature mode of "a fantastic scenario under the authority of a historian."

The most devastating point is self-referential. The lecture itself is an example of the very thing it warns against: a coherent, flattering, alarming, simplifying story optimized for circulation from the stage. And by his own criterion, there is no way to distinguish his grand synthesis from a powerful fiction that simply "connects" the audience well. He does not provide the external testing of his totalizing claims that, by his own theory, a healthy information network requires. Coherence is not truth; and his entire construction rests precisely on coherence.

## Conclusion

I am sick with revulsion toward Harari. This is personal, as is the insult in the article's title. This is a personal blog, and here I amuse myself however I like. But that revulsion should be kept precise. The key point is not that "he talks nonsense," but the form: totalizing confidence, unfalsifiable scale, self-flattering simplification, and speculative inflation. He uses exactly the technique he warns against, and he fails his own test for truth.

---

#### Sources

- On Harari having "slipped past fact-checking by asking too-large questions" (the words of his doctoral supervisor, S. Gunn), and for an overview of academic criticism: [Wikipedia, Yuval Noah Harari](https://en.wikipedia.org/wiki/Yuval_Noah_Harari); ["A critique of Harari's popular-science writing", Why Evolution Is - C. Hallpike's review: [Hallpike, Review of *Sapiens* - Harari's "populist science": [The Dangerous Populist Science of Yuval Noah Harari, Current - The content of the talk: [Yuval Noah Harari: Why advanced societies fall for mass delusion, Big the line of argument is in his book [*Nexus*](https://www.ynharari.com/book/nexus/).