The Age of AI: And Our Human Future

What is this book? A treaty? A compendium? An impassioned plea? Or a sales pitch drawing its unique proposition from philosophy, history, and humanities littered with revelations of opportunities and doom?It is everything, and more.As of writing, GPT-4 is the rave in artificial intelligence (AI). But the world has already had intimations of AI's power through its predecessors, particularly ChatGPT. In a deluge of information, it is necessary to have a voice of authority and wisdom to explain the phenomenon we face. This book fits the bill as a pamphlet that conceptually explains what AI is, without the technicalities that may baffle non-technical readers. (I have enjoyed an interview between one of the authors, Eric Schmidt and Alexandr Wang, on the subject, which inspired my purchase of the book).As a compendium, the book catalogues the development in computing and situates modern AI as the culmination of years of progress. It posits that we have created a thing with processing power that outstrips human cognition and can capture aspects of reality beyond human detection. AI can now beat us, quite literally, at our own game, as seen in chess, where an AI trained on the rules of the game augments itself to make independent and more compelling moves beyond human comprehension. Similarly, AI can discover new antibiotics in record time by merely being exposed to fundamental principles. Chess and medical breakthroughs are frequent references in the book, demonstrating the extent to which AI would affect domains once reserved for humans.Chapter 2 is particularly delightful, with sentences brimming with such verve that one wishes it never ends. It explores centuries of sociocultural and sociopolitical forces that paved the way for AI and pranced through the evolution of human ingenuity, reason, and intellect. The chapter posits that AI's ability to upend every aspect of society surpasses the revolutions wrought by the printing press and electricity. These earlier technologies not only introduced new forms; they disrupted every aspect of society. The printing press bestowed new roles on the Western individual by wresting powers away from the Church and equipping the individual - facilitated by the Protestant agitations - with scholarly access to the divine. This psychological shift in the Western mind - sufficiently explored in Joseph Henrich's work, The Weirdest People in the World - launched the Renaissance, ushering in flourishing in arts, architecture, literature, and civic participation, ensuring the greatness of Michelangelo, Leonardo da Vinci, and others. The authors contend that AI is destined to follow a similar trajectory.The philosophical underpinnings of AI are captivating. The authors, probing AI's capture of reality, align it with Wittgenstein's view of making meaning through familiar connections rather than reducing reality to mechanistic explanations. Hence, the neural networks that inspire AI in mimicking the structure of the human brain place it far from mere computations of cause and effect, or garbage in, garbage out.“To enable machine learning, what mattered was the overlap between various representations of a thing, not its ideal — in philosophical terms, Wittgenstein, not Plato. The modern field of machine learning — of programs that learn through experience — was born.”Still, on philosophy, the book wonders if we are equipped to deal with our new fate. If AI can capture reality outside human conception, how do we retain our identity when perceptions would be determined by something beyond us? The authors concede that civilisation has been primarily created and sustained through the dynamics of Faith and Reason, and AI is designing a new form. It is a difficult notion to digest, since phenomena that thinkers and philosophers have grappled with, e.g., consciousness, divinity, nature/nurture, would become more challenging to comprehend.There is a lot of caution in the book. The authors warn that where nuclear weapon is the most dreadful of human arsenals, AI surpasses it by an order of magnitude. This apocalyptic view is further compounded by the difficulty of designing effective verification systems for a rather inscrutable technology. It is to wonder what we have gotten ourselves into. Human ingenuity has birthed a hybrid of saint and devil. Where nuclear weapon is under international regulations in which nations with nuclear capabilities are under the watchful eyes of post-WWII and post-Soviet accords, how do we police something so insanely hard to detect, easily distributed, and accessible? Nuclear deterrence has so far saved us from annihilation. How do we protect ourselves from something that possesses the capacity to "transform conventional, nuclear, and cyber weapons strategy"? This makes the book an entreaty, inviting governments, policy wonks, and military thinkers to convene and hash out red lines that would ensure responsible applications.For me, the positives outweigh the negatives if regulations are in place. And we must be careful to avoid stifling innovation under the guise of potential misuse. Moreover, as AI accelerates prosperity and instigates breakthroughs, how will it impact the global south? Will it leave a section of humanity behind while perpetuating historical patterns of economic inequities, a fact that Emad Mostaque of Stability AI has been vocal about? The book hints at it, but it would take a separate publication to articulate this concern.Overall, it is a delightful book written by those who should write about AI and society.

The first half is very instructive with respect to describing AI and AGI and placing their likely impact on humanity in perspective. I also think the geopolitical impact of AI is well presented. I thought the second half was a little repetitive with respect to several themes concerning a nuanced management of AI. Overall, I learned a lot from the book and recommend it to other readers interested in the subject.

The Age of AI: And Our Human Future is an incredible read that dives deep into how artificial intelligence is changing our world. What I loved about this book is how the authors combine their unique perspectives. Kissinger brings a historical and philosophical lens, Schmidt offers insights from his time as Google’s CEO, and Huttenlocher adds depth with his expertise in AI. Together, they tackle some big questions: How is AI shaping industries and global power? What happens to human decision-making when machines can think faster than we can? They explain things in a way that’s easy to follow, even if you’re not a tech expert.This book isn’t just about the technology. It’s also about the future of humanity. It forces you to think about how we adapt to living in a world where machines are becoming more intelligent and independent and makes you question your own existence.If you’re curious about the future, how AI is reshaping the world, and what it means for us as humans, this book is a must-read. Thoughtful, engaging, and surprisingly approachable—I highly recommend it!

This is yet another book on how to keep ever-more-powerful AI beneficial for humanity. Although it carefully and eloquently discusses many important issues, it is disappointingly lacking in concrete suggestions for what to do and not do. Instead, it's main suggestion seems to be "just put people like us in charge and trust us – and we'll figure it out for the rest of you". Indeed, a key message at the end of the book is that the US should grant great power to a new national AI commission containing people just like Kissinger, Schmidt and Huttenlocher, "from the highest levels of government, business and academia".This is a book review rather than political commentary, so it is up to the reader rather than me to assess the wisdom of letting leaders from Google and Facebook shape our AI legislation. However, from a purely literary point of view, the irony is jarring when the book discusses the risks of big tech undermining democracy and working ever closer with government. Here is a brief summary of the book.The first three chapters present a brief history of information technology and artificial intelligence and its societal impact. The authors briefly raise the possibility that artificial general intelligence (AGI) may one day reach human capability across the board, but don't discuss the oft-raised concerns that this may engender extreme inequality, an Orwellian surveillance state or human extinction.The 4th chapter tackles global network platforms. It provides an interesting review of ever-larger platforms such as Facebook, Google, Uber and TikTok and their increasingly geopolitical implications. It would have been interesting to see more discussions of current controversies, for examples arguments for and against monopoly-busting, and discussion of the tradeoffs between content moderation and anti-disinformation on one hand versus censorship and propaganda on the other.The 5th chapter covers military AI. It contains interesting historical parallels, to which Henry Kissinger presumably contributed. It highlights risks of arms races and inadvertent AI-fueled escalation, and states that "a sober effort at AI arms control is not at odds with national security; it is an attempt to ensure that security is pursued and achieved in the context of a human future". Yet it again lacks concrete suggestions – for example, it does not express support for an international ban on lethal autonomous weapons which decide whom to kill without human intervention, or mention that this position is supported by the International Committee of the Red Cross and enjoys broad support among AI researchers. They merely suggest that "governments of technologically advanced countries should explore the challenges of mutual restraint", which would presumably let any country or company legally produce whatever cheap lethal autonomous weapons they wish to sell.The 6th chapter discusses the implications of powerful future AI for human identity.Although it argues that "our emphasis may need to shift from the centrality of human reason to the centrality to human dignity and autonomy", it would have been interesting to see arguments for how this autonomy can be retained in a society where even more decisions are made by ever smarter machines. It claims that AI and humans will become equal partners in many areas, without explaining how this can remain true if AGI is attained and keeps improving. The authors argue that "AI intermediation that prevents misinformation and disinformation will be crucial", without clarifying what or who will decide what is true – a timely free-speech question given that Huttenlocher's MIT just made national news by canceling an astronomy lecturer for writing a Newsweek article unrelated to his research.As mentioned, the last chapter makes a very concrete recommendation: that the US create a powerful AI commission with members like the authors. The proposed commission sounds a lot like the recently announced NAIAC, but with added powers, and it will be interesting to see if this book helps any of them get on it. The book consistently portrays the greatest threats to democratic societies as coming from others: from China, from U.S. citizens spreading misinformation, etc. They don't say much about the oft-discussed threats posed by the power centers they themselves represent. This would have been interesting given Kissinger's track record of toppling democratically elected governments and Google's and MIT's track record in arbitrating free speech. The book begs a very interesting question: If there is to be a powerful AI commission, whom would *you* trust to lead it and look after your interests?