Building Trust in the Age of Disinformation – What Future Awaits Us? | Yuval Noah Harari

Yuval Noah Harari discusses trust, democracy, European technological dependence, and AI’s influence over language with Prof. Marijana Grbeša. The YouTube description(opens in a new tab) dates the conversation to October 2025 in Zagreb, organized by IMC & 24Sata; the video was published on February 26, 2026. The final part turns to Israel, Palestine, and Ukraine.

Information systems change whom people trust

• Intelligence does not guarantee good judgment. Harari distinguishes information from truth and intelligence from wisdom. His argument is that capable people can make poor decisions from bad information, so producing more information—or greater intelligence—does not by itself solve collective error. He extends that distinction to the possibility of highly capable but deluded AI. 01:00–04:00(opens in a new tab)
• Democracy depends on a workable public conversation. Harari describes communication technology as infrastructure for large-scale democratic participation. His diagnosis of the present crisis is that new information systems have made it harder to listen and agree on basic facts, rather than simply widening ideological differences. This is his historical and causal interpretation, not a comparative study presented in the talk. 09:37–15:28(opens in a new tab)
• Distrust of institutions can coexist with trust in algorithms. A person may reject journalists while accepting the feed that selects their news. Harari calls this a shift in trust, not its disappearance: algorithms increasingly mediate human relationships. Later, he uses news editing to explain the power involved—choosing what people discuss can shape public life without directly telling them what to think. 15:28–18:28(opens in a new tab), 26:15–28:08(opens in a new tab)
• European dependence is a sovereignty question in his account. Harari argues that Europe could become a more independent AI participant by combining its scientific, economic, and human resources. His phrase “AI immigrants” is an analogy for systems that could displace work, change culture, and remain controlled from abroad. It is a political argument about technological dependence, not a literal description of migration or a settled prediction. 21:34–25:19(opens in a new tab)

Language, authority, and human value

• The meaning of “thinking” changes the comparison. Harari describes AI’s language abilities as a challenge to his own work as an author. He distinguishes arranging words into logical structures from thinking as a felt, sensory experience. He expects AI to exceed humans at the former but does not attribute the latter to present AI. His comparison with thoughts observed during meditation is an introspective analogy, not an account of how models are implemented. 28:08–33:16(opens in a new tab)
• Financial cooperation could become unintelligible to humans. Harari describes money and investment as shared stories that persuade people to entrust resources to one another. He imagines AI creating financial instruments that let AI systems trade and cooperate with each other, while humans cannot understand the concepts behind those transactions. His concern is not just difficult terminology: those transactions could determine who gets hired or fired. This is a forecast, not an outcome demonstrated in the conversation. 33:16–36:07(opens in a new tab)
• An AI interpreter could compete with religious authorities. In Harari’s account, a sacred text cannot answer a reader’s question, so a priest or rabbi interprets it. An AI answering questions about that text gives believers another place to turn. He argues that this could shift authority toward AI where religion treats texts as its ultimate authority. His claims that AI could recall every relevant text and outperform human interpreters are part of that argument, not capabilities verified here; the prediction also depends on his text-centered view of religion. 35:18–39:56(opens in a new tab)
• Shared concern need not imply broader political agreement. Asked about efforts to halt advanced AI development, Harari says he sympathizes and has signed earlier petitions, while explicitly being unsure whether he signed the particular one mentioned. He sees potential cooperation across political differences around the value of human beings, rather than treating intelligence alone as the measure of worth. 39:56–41:59(opens in a new tab)

The conversation also addresses war

Harari argues that peace between Israelis and Palestinians remains possible and requires acknowledging both populations’ rights and suffering. On Ukraine, he presents resistance to territorial conquest as important to the wider international order and warns of consequences if annexation becomes normalized. These closing passages are political arguments, not AI analysis; the interview offers no confident forecast of when either conflict will end. 42:03–46:01(opens in a new tab), 47:03–49:56(opens in a new tab)