For book club in January we read The Age of AI: And Our Human Future by Eric Schmidt, Henry Kissinger, and Daniel Huttenkolcher. This book while recent (published at the end of ’21) is actually already a bit out-of-date given the rate of innovation in the domain of Artificial Intelligence, but still a good read. Everyone at Book Club enjoyed the book (actually more than I did I think!)
Pros: the book provides an excellent recap of AI’s progression through to generative artificial intelligence (including ChatGPT) in ways that feel prescient given where we are two years after its publication. The authors delve into the implications for society and governments, and make recommendations (not on specific policy, but on previous methodologies used to craft such doctrine) with the goal of ensuring that AI can continue to be a force for good, and not a case of further enmity and inequity between societies that can and cannot embrace the technology.
Cons: there was far more attention given to war-time implications of AI, and the risks of this technology without specific solution-oriented recommendations. There are many things being done in academia and at public companies to reduce bias and also to increase participation across the world of the use of AI for solving important problems from healthcare to environmental protection; this was barely glossed over in initial chapters, and that felt like a miss. Ethical AI, and its distribution is a matter of major research and industry engagement right now. The main focus of the book was more on the implications for nation states and how critical it was to begin bringing companies and countries together to align on an appropriate iteration of a deterrence doctrine. While I don’t disagree that policy makers need to include technologists and companies advancing the state of the art here, I felt that was a narrow and defensive focus for a book whose title at least has far greater implications.
My other critique of the book (shared by more folks than my feeling that it was too focused on national defense implications) was that this book is VERY Eurocentric. All of the examples from the Renaissance to the research are centered in Europe and America. While there is discussion about China and its specific national investments, no other nation outside of America and Europe is specifically mentioned (except Iran as a historical example of how nuclear policies unfolded). If one truly wants to give a call to action about the role of AI in international policy, it would have been better to be more inclusive in one’s world view.
In general I liked the book, but I have found classes and books on LLMs, and the advancements in gene therapy, protein folding, etc. so much more interesting than the national security and defense implications. I was in the minority of our book club with that perspective, which is more likely a testament to my pacifism than a real critique of the book. It is exactly what I would have pictured coming from a Kissinger book. Definitely worth a read for the historical recap if nothing else.