In 2021, a trio of luminaries published The Age of AI and Our Human Future. As readings go, it felt worthy, but it wasn’t exciting. Anyone following the developments and uses for artificial intelligence already grasped the essentials.
Those authors were Henry Kissinger, former Google CEO Eric Schmidt, and MIT dean Daniel Huttenlocher. Their book focused on philosophical questions and the development of AI. For something meant to be a big deal, though, their approach was cautious. It’s good to have a conservative attitude towards developments so fast-moving, but The Age of AI lacked imagination.
Now, however, another Kissinger-led trio has published Genesis: Artificial Intelligence, Hope, and the Human Spirit. This time the book embraces the use of imagination—and makes for an engaging read. (It’s also short and the entire book can be read in a few hours.)
The authors’ willingness to ponder many ‘what-ifs’ testifies to the rapid changes in the field of AI in the three years since the previous book. ChatGPT went public in November 2022 and hit like a grenade.
Big questions not business models
This is Kissinger’s final book, published posthumously. Schmidt is again along for the ride, with Craig Mundie, senior advisor to Microsoft.
With former bigwigs from Google and Microsoft as co-authors, Genesis does not address the current business issues for American Big Tech, and the commercial viability of the vast amounts of money being poured into GenAI. Nor does it have anything to say about ‘AI slop’, the mass theft of IP, or the way the push for GenAI has corrupted Big Tech (try doing a search on Google these days, or dealing with the AI ‘content’ on Microsoft-owned LinkedIn).
Genesis instead remains fixed on the big questions. Its greatest use is speculating about the impact of superintelligence, whose advent it takes as a given. Usually a business book should be less about sci-fi and more about practicalities, but in this case, stretching the imagination is valuable. We’re all rushing around a corner; it is useful for entrepreneurs, executives, board directors, regulators, and investors to think about these ‘what-ifs’. These may cohere into practical questions all too soon.
The authors’ questions begin and end in the abstract, but they involve how and whether humans can keep control over AI, especially once the AI exceed our intelligence. From the opening:
“Do we pick our objectives and harness AI to achieve them, or do we let AIs help pick some of the objectives themselves?…That will require the resolution of not one but two ‘alignment’ problems: the technical alignment of human values and intentions with the actions of AI, and the diplomatic alignment of humans with their fellow humans.”
And from the conclusion:
“Might AI cause these coming crises [war, climate, etc], and then act as our savior—manufacturing problems that only it can solve, if only to prove its necessity and to remind us of our dependency? Again, we return to the dilemma that has motivated much of this volume: the excruciating choice between control and utility, between the comfort of the historically independent human and the possibilities of an entirely new partnership…
“A further unsolved question is Who will decide? Who will make the choice to delegate, or not delegate, responsibility and authority? Who will give, or withhold, resources? How will any one set of deciders communicate, converge—or come into conflict with—others attempting to make the same decisions elsewhere? Are we choosing those individuals, those fallible humans, now? Have we, unwittingly, already chosen?”
Control versus convenience
Underlying these questions is the observation that humans can’t understand how AIs derive their decisions: the ‘explainability’ problem. While businesspeople have been grappling with the practicalities of this for more than a decade, the authors contextualize it as a deeper challenge: that machines now threaten the human monopoly on grasping reality.
Moreover, the idea of the scientific method is under threat, if people simply defer to machine authority without questioning it—giving AIs the ‘mandate of heaven’ or a divine right. And what happens when machines embrace this authority, when they believe in it?
The authors strike a balance between grave concern and hope for what super-charged intelligence can do for us. Their interest is not to argue for dystopia or utopia but for us to consider how to thread that needle.
For example, “AI could also introduce new dynamics to financial markets and economic policies. It is easy to imagine, abstractly and in principle, how AI might go about creating wealth—even with access only to the internet. But even if we can invent currencies, systems, markets, and policies that respond wisely to the emergence of this scale of potential value-creation, how, exactly, would AI solve or end poverty, for example? As a practical matter, how would it establish an absolute global baseline for our quality of life?”
DigFin recommends Genesis as a thoughtful guide to the questions that businesspeople, financiers and entrepreneurs are just beginning to address. The heavyweight nature of its authorship assures a careful approach instead of the histrionics often associated with this topic. The authors are, nonetheless, bold in their questioning.
Genesis: Artificial Intelligence, Hope, and the Human Spirit”, by Henry Kissinger, Craig Mundie and Eric Schmidt; Little, Brown & Co., New York, 2024.
