Dr. John W. Sutherlin is a Professor of Political Science and Chief Innovation and Research Officer at the University of Louisiana Monroe.
To fully appreciate The Age of AI: And Our Human Future, one must overlook its nebulous description of a decades-old issue and suspend any expectations for a well-researched and thorough account of this vital topic. The authors, who represent major policy, industry, and academic heavyweights, stumble in their attempt to raise awareness and often fail to provide meaningful insights. The analysis and research manifested here leave so many things unanswered. In the end, many will ask themselves why they selected this book out of the choices currently available. This is not a typical Kissinger work spanning 800 or more pages with thousands of sources and infinitely quotable passages exhibiting personal perceptions and a vast foreign policy knowledge. Further, this is not a Schmidt work of pithy industry-level expertise with keen insights or observations about Google software packages. Perhaps the authors were less interested in an exhaustive treatment of artificial intelligence (AI) and more captivated with making a simple declaration, a clarion call to arms. However, even with this notion as the focal point, the reader may be left wanting more. Still, the book is not without merit; some may find it a good starting point for a deeper dive into the subject of AI and public policy.
Each chapter begins much the same, as philosophers and authors of antiquity are used to lay a foundation for banal statements regarding policy concerns about machines making human decisions. Descartes, Spinoza, and, of course, Kant are paraded before the reader, creating intellectual mediocrity and a confusion of cerebral demands. Perhaps these authors really believe that St. Thomas Aquinas and TikTok can elevate our ethical discourse. Maybe there will be some readers that find incorporating Clausewitz and Gutenberg into the tussle is essential. I did not find it to be particularly helpful.
If the authors want readers to think about the postmodern world where computers make decisions, then why revisit the Middle Ages? What the readers get are often ambiguous or obvious statements. This book would have been more relevant if it had been written 15 years ago. “AIs chiefly use data to perform tasks such as discovering trends, identifying images, and making predictions.” And? Does the process of shifting from physical maps to “network platforms using algorithms” really represent a paradigm shift that requires another book to document the eroding of human values and input? Hardly.
This effort falls somewhere between a book and a journal article. If the reader thinks of The Age of AI as a policy briefing, then most frustrations, disappointments, and regrets will vanish. The book is worth reading if for only one set of questions asked: “Are humans and AI approaching the same reality from different standpoints, with complementary strengths? Or do we perceive two different, partially overlapping realities: one that humans can elaborate through reason and another that AI can elaborate through algorithms?” Regardless of the policy area—that is, national security, health care, or commercial interactions—AI is still growing fast, with few human restraints and little thought about its potential repercussions for moral decisionmaking.
The authors insist that “governments, universities, and private-sector innovators should aim to establish limits.” I guess the question is “How?” AI has already proved it can beat the socks off human chess players. Is it too late to install safeguards that prevent AI from making fatal decisions where humans are the means to a silicon end? The authors point out that Alan Turing showed acumen in the 1950s and that GPT-3 (third generation generative pre-trained transformer) technology today is closely approaching what AI would define as “consciousness.” What is next? Algorithms fashioning popular music for us to purchase? AI making cost-benefit analysis for rationing medicine? Or deciding which cities to bomb?
Oops! Too late. The authors, correctly, find that the AI Rubicon has been crossed.
AI “permits us to aggregate and analyze data” more quickly and without any messy human emotions and biased reasoning. But this also means no human morals and ethics. This could have been the place for the discussion to begin about our human future. The authors ask us to consider an ethical construct as “paramount,” allowing political leaders an opportunity to engage with humanity. Without sufficient human (or governmental) limits, nations may simply default to AI for, inter alia, national policy decisionmaking.
Yet I wonder. What would happen to the nation that forwent its reliance on high-speed computers that evolve into AI, instead embracing human fallibility and the sluggish analysis of complex data? Would anyone burn the calculators in favor of the abacus? The need for humans to incorporate ethics into their tools has been around since at least Galileo.
The AI ship has sailed. Now, humans must constantly integrate their flawed beliefs into both social and silicon systems. AI consciousness may be only another terabyte away, so the authors are correct there. GPT-3, for example, lacks the ability to act independently . . . for now.
A better analysis on artificial intelligence and political power is Michael Kanaan’s book T-Minus AI: Humanity’s Countdown to Artificial Intelligence and the New Pursuit of Global Power (BenBella Books, 2020). Schmidt even praises Kanaan’s work as an excellent source of analysis. For those more interested in the nexus between AI and the military, Christian Brose’s The Kill Chain: Defending America in the Future of High-Tech Warfare (Hachette Books, 2020) is a better researched call to arms. JFQ