Book Review: SuperAgency (Hoffman and Beato)
What Could Possibly Go Right?!
“Even among AI developers, some believe that future instances of superintelligent AIs could represent an extinction level threat to humanity… As hard as it may be to accurately predict the future, it’s even harder to stop it… That’s because as much as collaboration defines us, competition does too… and that means you’ll never get the future you want simply by prohibiting the future you don’t want.” — SuperAgency, p. 12
This sobering yet hopeful opening sets the tone for SuperAgency, a timely and techno-humanist exploration of artificial intelligence and its role in shaping the future. Written by Reid Hoffman, co-founder of LinkedIn and investor at Greylock, and tech journalist Greg Beato, the book advances a provocative argument: AI will either amplify human agency—or undermine it entirely. It’s up to us to decide, but timing is critical.
You would be interested in this book if…
-
You’re trying to make sense of the current moment in AI development and where it’s headed.
-
You believe in the potential of AI to augment—not replace—human capabilities.
-
You want a clearer vocabulary around competing narratives: doomers, gloomers, zoomers, and bloomers.
-
You’re a school leader, policymaker, or educator wondering how to integrate digital fluency into your graduate profile.
-
You see the value in iterative deployment and agile innovation—but still want space for critique.
You would not enjoy this book if you’re looking for a deeply critical or wellness-centered perspective on AI. The authors’ optimism often overshadows concerns about digital wellbeing, power asymmetries, and ethical slow-walking.
A Techno-Humanist Compass
Hoffman and Beato channel McLuhan throughout the book, most clearly when they write:
“We continuously create new tools to amplify our capabilities and shape the world to our liking. In turn, these tools end up shaping us as well… Every new technology we’ve invented, from language to books to the mobile phone, has defined, redefined, deepened, and expanded what it means to be human.” (p. 14)
They call this perspective “techno-humanism”—the belief that humans and tools co-evolve and are integrated into our human development – always have been. The central question is no longer if AI will shape the world, but how we’ll shape AI so it reflects our values. Their thesis: when used well, AI can produce a new kind of collective lift—what they call superagency:
“Superagency is what happens when a critical mass of individuals, personally empowered by artificial intelligence, begin to operate at levels that compound through society.” (p. 13)
GenAI has the potential to level-up humanity, what we do well, and how we can still grow our humanity through an ethical and values-based lens. But it takes more than government to do this – the public must be involved. But how?
Iterative Deployment, Not Perfection
The authors push back against the precautionary principle that says AI should be deemed “guilty until proven innocent.” Instead, they argue for iterative deployment—a kind of real-time, collective stress test for AI systems:
“Instead of settling for nothing less than risk-free models, we should make it our goal to understand the risks that occur in real-world conditions and systematically work to manage and reduce them.” (p. 133)
They draw analogies to the automobile: laws, norms, and infrastructure emerged gradually as people started driving. Likewise, real-world AI use will inform better policy—if we’re proactive, engaged, and values-driven.
But the authors do glance over the many deaths that lead to these regulations. Much like they glance over the tragic consequences of GenAI use by youth, and naive users.
Competing Narratives: Doom, Gloom, Zoom, and Bloom
One of the book’s most useful frameworks is its categorization of public discourse on AI:
-
Doomers fear extinction-level threats from misaligned superintelligence.
-
Gloomers critique both AI and doomer hype, focusing instead on short-term harms like disinformation and job loss.
-
Zoomers advocate rapid, unregulated innovation to maximize productivity.
-
Bloomers, where the authors reside, are cautiously optimistic and want broad democratic participation in shaping AI’s trajectory.
In this way, SuperAgency aligns closely with civic technologists and education leaders who see AI as a tool for human flourishing, not just economic output. It might be possible to recognize yourself in many of these different definitions depending on your use; however, the authors do not allow for this more nuanced argument.
The Data Deluge and the Case for AI
On page 46, a staggering data point supports their urgency:
“The average U.S. inhabitant was already only consuming 0.004 of the information available each day in 1990. Today, in the time it takes you to read this sentence, the world produces enough data to fill 23 billion e-books.”
Their argument is simple: only AI has the capacity to synthesize this flood of information into actionable insight. In their equation, Big Data → Big Knowledge → Value Creation.
But the trade-off is real. They admit: “Phones are the first thing people reach for in the morning, the last thing they interact with at night.” (p. 96). Yet they fail to address what this means for digital well-being, cognitive overload, or the loss of solitude. Their model of agency is productivity-centered, not necessarily human-centered.
Why This Matters for Educators & School Leaders
“Life as a human today means constantly upskilling… The 21st century has given us AI.” (p. 149)
In this world, education must prepare students not just for tools of the future, but for decisions about the future. For schools like Kingsway College School, this book resonates with our Portrait of a Graduate and our focus on student agency, ethics, and inquiry.
The authors argue that ethics, norms, and shared values—not just policies—must guide how we deploy AI:
“Citizens will play an active and substantive role in legitimizing AI.” (p. 184)
This reminds us that educators, students, and families have critical roles in shaping the discourse—not just tech firms and regulators. The book does a good job of starting with the basics of what LLMs are, and how best to use them, so that is good; however, I think a more in-depth look at the precautionary principle, and at cautionary tails would have made this book more well-rounded.
Final Thoughts: A Book for the Zoomers and Bloomers
This is an engaging, well-researched book with a clear message: we are still early in the AI journey. Optimism is warranted—but only if it comes with vigilance.
While SuperAgency offers a compelling case for iterative deployment and techno-humanism, it glosses over well-being, ethics of care, and the power imbalances baked into current AI architectures.
Still, I recommend this book—especially for those in the bloomer camp who want to champion human agency in a post-AI world. Read it not as gospel, but as a spark.