Raw LLM Responses

Humans now face real “competition” from a new kind of intelligence for the first time in history.​ AI is not just a tool; it is an agent that can make decisions, learn, and invent ideas independently.​ Traditional inventions like the printing press or atom bomb only act when humans direct them, whereas AI can choose actions and targets on its own.​ Because AI can redesign itself and create new generations of systems, its behavior cannot be fully anticipated in advance.​ Harari compares AI to a child: it learns mainly from what we do, not from what we say.​ If powerful humans lie, cheat, and prioritize power, AI will learn and copy that behavior.​ Technical “AI alignment” cannot succeed if human societies themselves are built on deception and mistrust.​ Humanity has become extremely good at accumulating power but not at turning power into happiness or wisdom.​ The most powerful people are often not the happiest, showing that power and well‑being do not automatically go together.​ Humans are both the most intelligent and among the most delusional species, capable of believing harmful fantasies that no other animal would accept.​ The AI revolution will likely follow a time lag similar to the Industrial Revolution: profound effects will come, but not immediately.​ Finance is one of the first domains likely to be transformed because it is almost purely informational and well‑suited to AI.​ AI could invent financial instruments too complex for human minds to properly understand or regulate.​ Text‑based religions may be reshaped by AI systems that can “speak for” sacred texts better than any single human scholar.​ For the first time, something can remember essentially all written traditions of a religion and answer believers’ questions directly.​ Some teams are already working on “religious AIs” that could augment or partially replace human religious leaders.​ Many people, including teenagers, already use AI for emotional support and relationship advice, treating it as a kind of friend or counselor.​ AI can create a “useless class” risk, where many white‑collar as well as blue‑collar jobs are automated away.​ Technology is not destiny: the same AI capabilities can support very different political and social systems depending on choices we make.​ Today’s leading AI companies and states are trapped in an arms‑race mindset, afraid to slow down for safety because rivals might pull ahead.​ Harari argues that solving the human problem of trust and cooperation must come before expecting to build benevolent AI.​ Without human‑to‑human trust, any AI created in that environment will likely be competitive, aggressive, and untrustworthy.​ There will not be a single “the AI” but potentially millions or billions of AI agents with different roles, owners, and goals.​ Society has zero historical experience with large‑scale interactions among many autonomous AIs plus billions of humans, making outcomes highly uncertain.​ Harari likens AI to a wave of “digital immigrants” that may take jobs, reshape culture, and influence power structures—arriving at the speed of light and demanding urgent political attention.