Yuval Noah Harari: Free Speech, Institutional Distrust, & Social Order | Making Sense #386
Sam Harris and Yuval Noah Harari discuss Harari’s book Nexus and its core question: why technologically capable humans continue making collectively dangerous...
43mSummary published by 1% Better, updated .
Key Takeaway
Treat your information intake like a diet today: before sharing or trusting a claim, pause to ask what evidence supports it, who is accountable for it, and whether it has been amplified by an algorithm rather than earned attention. Harari’s central point is that truth is costly, complex, and sometimes uncomfortable, while fiction spreads cheaply. Protect your judgment by favoring accountable institutions, diverse checks, and correction over outrage or total distrust.
Episode Overview
Sam Harris and Yuval Noah Harari discuss Harari’s book Nexus and its core question: why technologically capable humans continue making collectively dangerous decisions. They examine information networks, institutional trust, democracy’s self-correcting mechanisms, the tension between truth and social order, and how social-media algorithms and bot accounts distort public conversation.
Main Insights
1. Put Your Information on a Diet
Harari argues that more information is not inherently better, comparing unrestricted information consumption to overeating junk food. Because truth is costly to produce while fiction is cheap, simple, and often emotionally appealing, individuals need to be more selective about what they consume, trust, and amplify.
2. Repair Institutions Instead of Abandoning Them
Institutions can be mistaken, captured, or corrupted because they are made of fallible people, Harari says. But the answer is multiple institutions that check one another and internal mechanisms for identifying and correcting errors—not wholesale destruction of the systems that make collective life possible.
3. Distinguish a Claim’s Existence From Its Amplification
Harari recommends being highly cautious about censoring human users, even when they publish falsehoods. His sharper line is that platforms should be accountable when their own algorithms choose to promote a false claim, because amplification is an editorial decision rather than merely the passive hosting of speech.
4. Treat Algorithmic Curation as Editorial Power
Harari notes that algorithms have effectively automated the role once held by news editors: deciding what people see and discuss first. Since this placement shapes public attention, he argues platforms should bear responsibility for the consequences of their recommendation systems.
5. Demand Bot Disclosure Before Trusting Social Proof
Bots may participate in useful ways, including potential medical applications, but Harari says they should never masquerade as people. When artificial accounts create the appearance that many humans care about a topic, they can manipulate what real people believe deserves attention.
6. Use Self-Correction as the Test of a Healthy System
Harari identifies self-correction as the defining feature of a good institution. Elections, free media, courts, and other independent bodies provide ways to expose errors and change course; dictatorships fail this test because there is no reliable mechanism to correct the leader’s mistakes.
7. Reject the Reflex That Everything Is Power
Harari warns that an extreme cynicism shared across parts of the political left and right reduces every institution and claim to a power struggle. He argues this view destroys the trust democracy needs and overlooks the fact that at least some people—including powerful people—genuinely care about truth.
8. Hold Conventions Lightly, Not Recklessly
Large societies depend on shared conventions, such as money and rules of coordination, that are human-made rather than objective facts. Harari’s distinction is not between using or rejecting such fictions, but between treating them as amendable human creations and treating them as beyond revision.
9. Balance Truth-Seeking With Social Coordination
Information networks must produce both truth and order, according to Harari. Facts are essential for projects such as building an atom bomb, but facts alone do not organize millions of people; ideology and shared stories coordinate action, creating an ongoing trade-off that requires scrutiny.
10. Choose the Middle Path Through Ongoing Democratic Debate
Harari rejects extreme, all-or-nothing positions on surveillance, immigration, and free speech. The practical democratic posture is to assume fallibility, hear competing views, test policies, and keep correcting mistakes rather than treating every disagreement as a war between good and evil.
Frameworks or Models
Democratic Self-Correction
- Distribute authority across several institutions rather than relying on one center. 2. Give institutions mechanisms to expose and correct their own errors. 3. Use elections, free media, and courts to challenge failures and change course. 4. Continue correcting because each safeguard is itself fallible.
Accountable Social-Media Governance
- Preserve human users’ ability to speak, including when they are wrong, with caution around censorship. 2. Hold companies liable when their algorithms amplify harmful or false content. 3. Require bots and algorithms to identify themselves rather than impersonating humans. 4. Apply a fiduciary-style duty to platforms’ use of users’ private information.
Notable Quotes
"if you give good people bad information they make bad decisions it's it's as simple as that"
"the world basically needs an information diet and you know the truth is a subset of information and a very small subset"
"this is the sign of a good Institution that it has mechanisms inside the institution to identify and correct its own mistakes"
"democracy is a conversation imagine it as a group of people standing in a circle talking with each other suddenly a group of robots join the circle and talk very loudly very persuasively they pretend to be humans and you don't know you can't tell who are the humans and who are the robots so the conversation breaks down"
Action Items
Suggested exercises inspired by the conversation.
-
1
Add a verification pause
Before reposting a provocative claim today, spend five minutes finding its original source, checking whether a credible institution has evaluated it, and separating evidence from emotionally satisfying narrative.
-
2
Audit your algorithmic inputs
Review several accounts, channels, or feeds you follow. Unfollow or mute sources that repeatedly optimize for outrage, certainty, or unverified claims, and add at least one accountable source with transparent editorial standards.
-
3
Label social proof skeptically
When a topic appears to be trending, ask whether its apparent popularity could be algorithmic or bot-driven. Do not infer importance, consensus, or truth merely from engagement counts.
-
4
Practice corrective disagreement
In one political or workplace disagreement this week, state a position as provisional, ask what evidence would change your mind, and identify one part of the other view that may be reasonable.
Guides that reference this episode
See how this episode fits with recommendations from other leading podcasts.
Go deeper
Full Transcript
Read or copy the complete episode transcript.