The Coming Wave

The Coming Wave is a book that cautions us to prepare for AI's disruption of our way of life. Mustafa Suleyman, the co-founder of DeepMind, presents a persuasive argument on the perils and promises of AI. He also highlights the promise of AI in concert with synthetic biology and quantum computing. His core argument is that a technological wave is coming, and we are underprepared.

The book is written in four parts. In the first part, Homo Technologicus, he establishes a correlation between human evolution and the advent of technology. He argues that human civilisation can be envisioned as one technological jump to another. Technology eventually permeates our way of life, so much so that it is eventually in our subconsciousness. For instance, electricity disrupted the world about a hundred years ago; today, we hardly think about it. Suleyman enunciates the concept of containment, which limits the spread of technology, and posits that the only technology that has been effectively contained is nuclear. He concludes that technology cannot be contained as "civilisations appetite for useful and cheaper technology is boundless."

In the second part, The Next Wave, he describes the advent of AI and its immense possibilities. The first stage of AI's evolution was "classification and prediction". He argues that this stage is behind us, and AI is becoming increasingly capable; the following stages are the development of artificial capable intelligence (ACI) and artificial general intelligence (AGI). He informs the reader that the threat from AI is "deeper and more powerful than just another technology." He describes the dangerous cocktail of bio-synthesis and AI, highlighting the threats of gene modification. Suleyman suggests that coming technology waves are powered by clusters of technology peaking simultaneously. The cluster comprises AI, advanced biotechnology, quantum computing and robotics. The technology wave will be asymmetric, fast, and omni-use, giving a degree of autonomy of the kind not seen before. Eventually, "Homotechnologicus may be threatened by his own creation."

The third part of the book, States of Failure, outlines the challenges in regulating the technological wave. The wave will shock society, and we will be unprepared. He cautions that technology will also be a tool of power wielded by the powerful few, and the current governmental safeguards are inadequate. He warns that the charge of AI is also the beginning of an era of pervasive surveillance. A powerful AI tool could have the potential to overwrite its programming, even discarding the safety mechanisms built to regulate it.

In the book's final part, he proposes solutions to manage AI with the bottom line that technology has to be regulated to the extent that it does more good than harm. He suggests a ten-step approach towards containment. The plan could have technical safety guidelines and audit mechanisms, control the pace of technological surge by building time and procedure-based choke points, develop international guidelines and mitigation measures, and nurture a cooperative regulatory model between the innovators and governments. Technologies can evolve in ways unimaginable with far-reaching consequences. AI regulation will be possible if humanity can create the proper context and develop suitable mechanisms.

What works for the book is the author's storytelling flair. His narrative is vibrant, peppered with anecdotes, insights into technological landmarks, theoretical analysis, and thought-provoking questions for the reader. He handles a technical subject by dumbing it down so that the book is almost jargon-free. His ability to simplify complex processes is also noteworthy.

What does not work is the ponderous examination of the notion of containment. Suleyman spends far too many words to belabour the requirement of containment and the challenges in getting it right. He appears to have surrendered depth for width. The book skims the surfaces and presents a fundamental view of the challenges of AI but does not delve deep enough to find implementable solutions. A further downside to his book is that he does not offer a counternarrative. Drawing on industry experts and academics, he could have provided a more detailed counterview to his central arguments.

Overall, if you want to understand how the industry views challenges from AI and what the future portends, it's a perfect book to begin with. The book is a notable addition to the genre that explores the challenges emanating from the rise of AI. Suleyman's book is ambitious and sits at the junction of governance, technology management and humanity's future. If you want a book that offers a basic understanding of AI's challenges and a broad overview of the impending technological wave, go for it. For a deeper understanding of AI, maybe the answers lie elsewhere.

My Rating 3.5/5