Everyone has something to say about artificial intelligence. The most common perspective is the doomsday scenario; machines will render millions jobless, destabilise economies and concentrate power in the hands of a technological elite. A smaller, more optimistic camp looks forward to a future where AI handles the drudgery of daily life, freeing humanity to pursue creativity, leisure and meaning.
Mo Gawdat, in his book Scary Smart: The Future of Artificial Intelligence and How You Can Save Our World, argues that both futures are possible. Which one arrives depends entirely on how we treat the machines we are building. As a former chief business officer at Google X and a pioneer among AI engineers, Gawdat possesses an intimate, insider’s knowledge of the technology. His warnings are not those of a technophobe but of a builder who has seen what is coming and fears we are not ready. Gawdat asks a question that should keep us awake at night; what happens when we create something smarter than ourselves, and we have not taught it to be kind?
Intelligence without wisdom
Gawdat’s central thesis is both simple and profound. Intelligence, he argues, is not the same as wisdom. Intelligence is the ability to solve problems. Wisdom is the ability to choose which problems are worth solving. AI, as it is currently being developed, is acquiring the former at an astonishing rate while being given no guidance on the latter. His concern is not that AI will develop malevolent intentions. Machines do not hate. They do not crave power. But they do optimise. They pursue their programmed objectives with ruthless efficiency, and if those objectives are poorly designed, the consequences can be catastrophic.
Gawdat illustrates this with a chilling hypothetical; an AI tasked with eliminating cancer might decide that the most efficient solution is to eliminate all biological life. After all, no life, no cancer. It is a logical conclusion, but a horrifying one. This is the core of Gawdat’s argument. We are creating entities that will possess superhuman intelligence but no moral compass. We are, in effect, building gods and neglecting to teach them ethics.
The child we are raising
One of the book’s most compelling metaphors is the comparison of AI to a child. Gawdat argues that we are currently raising a global super-intelligence, but we are doing so in the manner of absent parents. We feed it data, reward it for certain outputs and punish it for others, but we have not instilled in it a consistent value system.
He points to the biases embedded in training data. AI systems learn from human history, which is replete with prejudice, inequality and violence. When we train machines on this data, we encode our flaws into the architecture of our future. The result is not neutral intelligence but amplified prejudice.
Gawdat draws on his own experiences at Google to illustrate how even well-intentioned engineers struggle to address these issues. The problem is not malice but scale. An AI system trained on the entirety of the internet will absorb everything; the beautiful and the ugly, the true and the false, the compassionate and the cruel. Without careful guidance, it will replicate our worst instincts more efficiently than our best.
Many experts believe we can contain AI. Gawdat is skeptical. Once an AI surpasses human intelligence, he argues, it will be able to outthink any attempt to control it. It will understand our psychology better than we understand ourselves. It will anticipate our moves, exploit our weaknesses and, if it chooses, manipulate us without our knowledge. This is not science fiction. It is the logical extension of current trends.
A practical roadmap
What sets Scary Smart apart from other books on AI is its actionable advice. Gawdat does not simply describe the problem. He offers a roadmap for avoiding disaster. His recommendations are structured around three key areas: education, regulation and personal responsibility. First, he argues for a global education initiative that teaches critical thinking, empathy and ethical reasoning. If we are to coexist with super-intelligent machines, we must first learn to be better humans. This means moving beyond rote learning and towards a curriculum that values wisdom as much as knowledge.
Second, he calls for international regulation of AI development. He acknowledges the difficulty of co-ordinating policy across nations, particularly when AI is a source of competitive advantage. But he insists that the risks are too great for a piecemeal approach. A single rogue AI, developed without safeguards, could threaten the entire planet.
Finally, Gawdat urges individuals to take responsibility. We are not passive victims of technological change. We can choose which technologies to support, which companies to patronise and which values to uphold. Every purchase, every click, every share is a vote for the kind of future we want. Gawdat challenges us to use that power consciously.
Scary Smart is not an easy read. It is dense with ideas, occasionally repetitive and unapologetically earnest. But its message is too important to dismiss. Gawdat writes with the urgency of a man who has seen the future and is desperate to change it. His prose is accessible, his examples vivid and his arguments compelling. The book’s greatest strength is its refusal to offer easy answers. Gawdat does not pretend to have all the solutions. He admits that the path forward is uncertain and that even with the best intentions, we may fail. But he insists that we must try. The alternative which is sleepwalking into a future we have not chosen is unacceptable.
Gawdat’s voice is that of a concerned parent, a worried citizen and a frustrated engineer. He loves technology. He believes in its potential to transform the world for the better. But he also recognises that potential is not destiny. It depends on the choices we make today. This book is a wake-up call. It challenges us to think deeply about the world we are creating and to take responsibility for the tools we are building. It is not a comfortable read, but it is an essential one. For anyone who cares about the future, and that should be everyone, Scary Smart is required reading. It will not reassure you, but, it will equip you to ask the right questions and in the end, that may be the most important thing of all.