🧠 Yuval Noah Harari Warns: AI May Be Humanity’s First Real Rival

Historian Yuval Noah Harari, author of Sapiens and Homo Deus, recently spoke with The Wall Street Journal about the transformative—and potentially dangerous—rise of artificial intelligence. Here are the key takeaways:

• AI Means Competition for Homo sapiens: For the first time in history, humans face real competition from non-human intelligence. AI agents can learn, evolve, and make decisions, representing an evolutionary turning point.
• The “Digital Immigration” Wave: Harari compares AI’s rise to a digital immigration event. Billions of digital agents—each with their own languages, goals, and economic functions—are flooding into human systems. Like immigration, this poses cultural and structural disruptions.
• Decentralized AIs—No Unified Authority: There won’t be one centralized AI. Instead, a competitive ecosystem of AIs created by corporations, governments, and individuals will emerge, each with its own influence and worldview.
• AI Is Not Just a Tool—It’s an Agent: Unlike past technologies, AI acts independently. It doesn’t just extend human ability—it generates new behaviors, decisions, and possibilities outside human control.
• The Pitfalls of AI “Education”: Just like children, AIs learn from what we do—not just what we say. If humans lie, cheat, and mislead, AI will mirror that. Ethics can’t be programmed in isolation—they must be modeled.

🔍 Reflections: Preparing Our Children to Compete with AI

As I reflect on Yuval Noah Harari’s warnings about AI, one truth hits hard: my day job and the other jobs we do today likely won’t exist for my children. Much like blacksmiths, switchboard operators, or seamstresses of the early 1900s, many of today’s professions will fade or fundamentally change. The pace of change, driven by artificial intelligence, is exponential (although it will follow the typical S-shape adoption curve).

So how do I equip my children for a future that doesn’t yet exist? I’m convinced that many of the jobs they’ll hold haven’t even been invented. While we’ll always need plumbers and people who can fix physical things in the real world, the vast majority of white-collar jobs are already being reimagined by AI—and will continue to evolve beyond recognition.

The skills they’ll need will go far beyond what traditional school systems are designed to teach. Will they need the ability to read social, emotional, and contextual cues in human-AI interaction. Will they need to know how to work with AI—prompting it skillfully, recognizing its blind spots, and demanding honest, accurate answers. Will they need to know to recognize when it hallucinates but also when it knows the truth and is lying? Research itself may be less about Googling and more about conducting AI-augmented investigations. And coding? It’s shifting too—from writing syntax to designing workflows and behaviors through that newly fangled computer language called English.

And yet AI will take a while to ascend into godhood. Projects like the LLM speedrunner challenge (conclusion paper) show that recursive self-improvement—AI that upgrades itself into something smarter over and over again—is not a reality yet. But that “yet” matters. What stage of AI recursive self-improvement will we be at by the time my kids are adults and in the workforce? What impact will that have on the way my children use and compete with AI?