This is a title that intrigues, at first glance, and which reflects the general tone of this work which seeks to demolish the ideas received on AI and to stimulate our intellect to take advantage of it in a reasoned and reasonable use, bringing great progress in all areas of social life, scientific life and business life. The title “anti-manual” means that there is not one AI but AI with different expectations and models for each of them. There is, therefore, no perfect user manual, but a progressive learning, both empirical and scientific, of the immense, even unlimited possibilities offered by these technologies.
In the broadest sense, AI encompasses all the techniques that can replicate the cognitive behavior of the human brain on machines. Beyond a simple replication, the explosion of absorption capacities, by machines of increasingly large volumes of data, has made it possible to develop uses of these increasingly promising technologies. This is the case of “machine learning” which, in a very simplistic way, beyond the automation of existing processes, will strive to optimize the use of computers based on real examples that will make it adapt new ways of working, for example, in the medical, legal, anti-fraud fields… Beyond the restructuring of complex processes, the potential of AI extends to the powerful creation of new organizational models in all areas of society, “deep learning”, consuming exponentially growing volumes of data. If the potential for progress is considerable, the ethical and geopolitical risks are immense and can threaten individual freedom.
In short, there is a risk for our democracies. One requirement is the implementation of very strict governance rules at both the national and global levels. One of the strengths of the book is its clarity and the many examples and anecdotes that illustrate the point and fortunately complement a very broad, very rich and sometimes difficult subject. This enlightening book deserves a prize.
Denis Molho