I started reading A BILLION TIMES SMARTER by Ingo Paas, and one thing became immediately clear: this is not another “AI productivity” book. It shares a human look on power, autonomy, and the changing relationship between people and intelligence itself.
Paas argues that humanity is entering a completely new era, the one where intelligence is no longer exclusively human because AI systems are already beginning to outperform people in speed, scale, memory, coordination, and increasingly, decision-making impact.
One of the strongest ideas in the book is this: “Power shifts once intelligence shapes existential impact at the systemic level.” That may sound dramatic, but when you look around, it already feels real. AI is now deeply integrated into: healthcare, software development, logistics, education, financial systems, cybersecurity, scientific discovery, and increasingly, daily personal decisions.
According to McKinsey & Company, AI adoption in organizations continues growing rapidly every year, while generative AI tools have become part of everyday workflows across industries. At the same time, studies from Stanford HAI show that AI model capabilities are advancing much faster than most governance systems or public understanding.
And honestly, this is probably the most important point: the debate is no longer about whether AI should exist, that discussion is over. Humanity has already entered the AI age, and the process cannot realistically be reversed. Everybody uses AI, sometimes consciously, sometimes invisibly through platforms, recommendations, search engines, copilots, autonomous systems, and digital assistants. And its use will only continue increasing exponentially over the next years.
Trying to “ban” or completely reject AI now would probably make as much sense as trying to stop the Internet in the 1990s. The benefits are obvious:
But the dangers are equally obvious: concentration of power, loss of human oversight, dependency on systems we no longer fully understand, autonomous decision-making, disinformation at massive scale, erosion of human agency. Paas describes this as the “AI Autonomy Paradox”, the same technology that can empower humanity may also gradually reduce human control over systems and decisions.
And to be fair, this concern is not unique to him. Researchers increasingly discuss the gap between AI performance and genuine understanding. A well-known academic paper, Artificial Intelligence “The Generative AI Paradox,” notes that modern models can produce expert-level outputs while still lacking deeper human-style comprehension.
This creates a strange new reality: AI can already outperform people in certain tasks while still behaving unpredictably in others. That combination is powerful and potentially dangerous. But at the same time, denying reality is not a strategy. For people who are genuinely trying to understand what this new technological era means, the most reasonable approach is probably not fear and not blind optimism either. It is responsibility (!) The real challenge now is not stopping AI, but keeping it under meaningful human control:
Interestingly, this idea is becoming more common even among AI practitioners themselves. One recent article on agentic AI described the future not as “humans replaced by AI,” but as “human-centered autonomy,” where AI systems increase human capability while humans remain responsible for direction and judgment. That distinction matters enormously because technology itself is never purely good or bad. What matters is who controls it, how responsibly it is deployed, and whether people remain conscious participants in the systems they create.
Paas repeatedly returns to one uncomfortable but important idea: humanity may not lose relevance because machines become “evil,” but because people gradually surrender too much decision-making to systems optimized for efficiency. And honestly, that may already be starting.
The AI era is here, the acceleration is real, the benefits are extraordinary, but the risks are real too. Now, the question is whether society can mature fast enough to handle the intelligence it has created.