What do you think?
Rate this book
The New Fire: War, Peace, and Democracy in the Age of AI
AI is revolutionizing the world. Here’s how democracies can come out on top.Artificial intelligence is revolutionizing the modern world. It is ubiquitous—in our homes and offices, in the present and most certainly in the future. Today, we encounter AI as our distant ancestors once encountered fire. If we manage AI well, it will become a force for good, lighting the way to many transformative inventions. If we deploy it thoughtlessly, it will advance beyond our control. If we wield it for destruction, it will fan the flames of a new kind of war, one that holds democracy in the balance. As AI policy experts Ben Buchanan and Andrew Imbrie show in The New Fire, few choices are more urgent—or more fascinating—than how we harness this technology and for what purpose. The new fire has three data, algorithms, and computing power. These components fuel viral disinformation campaigns, new hacking tools, and military weapons that once seemed like science fiction. To autocrats, AI offers the prospect of centralized control at home and asymmetric advantages in combat. It is easy to assume that democracies, bound by ethical constraints and disjointed in their approach, will be unable to keep up. But such a dystopia is hardly preordained. Combining an incisive understanding of technology with shrewd geopolitical analysis, Buchanan and Imbrie show how AI can work for democracy. With the right approach, technology need not favor tyranny.
338 pages, Kindle Edition
Published March 8, 2022
23 people are currently reading
506 people want to read
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 15 of 15 reviews
May 4, 2022
3.5 stars. A clear and accessible overview of developments in AI as a transformative technology. Many interesting details about automated hacking, semiconductor supply chains, great power dynamics, and regulatory scleroticism. And the final chapter contains an impressive list of concrete policy recommendations for democracies.
There wasn't a great deal of serious thinking about the prospect of truly transformative or 'general' AI, nor about related concerns around building safe and aligned AI. There is the familiar roster of concerns around social media algorithms boosting disinformation, bias in classifier models, political deepfakes. All clearly important and pressing problems, but not obviously the only challenges we may have to face if you really buy that "we encounter AI as our distant ancestors once encountered fire".
If you have not read books like Human Compatible or The Alignment Problem, I'd tentatively suggest starting there. If you have read them, much of the (very good) overviews in this book are likely skimmable.
Incidentally I really like that central analogy. Fire is powerful, but by default undirected. Learning enough to light fires, but not enough to control them, could mean burning down your home. When early humans learned how to light fires, they razed landscapes in ways that are still visible today. Fortunately, we're now able to use fire like a tool.
Imagine you are an early human living on a small forested island, unaware of life beyond your island. You have discovered fire, and so far managed small and short-lived demonstrations. As you learn to use it to cook and stay warm, you begin to worry about who should be able to create fire, and how to distribute the cooked food it produces. Then the weather turns hot and dry, and you notice the possibility of a forest fire: destroying the island that sustains you. This prospect seems worth worrying about at least as much, even though you've not witnessed anything really like it before. So you might consider setting rules for only lighting fires on the beach, or building firebreaks, and otherwise learning how to control fire before it burns down your home.
There wasn't a great deal of serious thinking about the prospect of truly transformative or 'general' AI, nor about related concerns around building safe and aligned AI. There is the familiar roster of concerns around social media algorithms boosting disinformation, bias in classifier models, political deepfakes. All clearly important and pressing problems, but not obviously the only challenges we may have to face if you really buy that "we encounter AI as our distant ancestors once encountered fire".
If you have not read books like Human Compatible or The Alignment Problem, I'd tentatively suggest starting there. If you have read them, much of the (very good) overviews in this book are likely skimmable.
Incidentally I really like that central analogy. Fire is powerful, but by default undirected. Learning enough to light fires, but not enough to control them, could mean burning down your home. When early humans learned how to light fires, they razed landscapes in ways that are still visible today. Fortunately, we're now able to use fire like a tool.
Imagine you are an early human living on a small forested island, unaware of life beyond your island. You have discovered fire, and so far managed small and short-lived demonstrations. As you learn to use it to cook and stay warm, you begin to worry about who should be able to create fire, and how to distribute the cooked food it produces. Then the weather turns hot and dry, and you notice the possibility of a forest fire: destroying the island that sustains you. This prospect seems worth worrying about at least as much, even though you've not witnessed anything really like it before. So you might consider setting rules for only lighting fires on the beach, or building firebreaks, and otherwise learning how to control fire before it burns down your home.
December 13, 2021
The New Fire is a fascinating book about artificial intelligence (AI) and what it means for the future of geopolitics. Written by two scholars of Georgetown University, it is remarkably accessible while diving deep into the subject matter. Complex concepts such as neural networks and generative adversarial networks (GAN) are explained in very clear terms, so that they can be understood by readers without a technical background. To make the book a more pleasant read, the authors also combine case studies with the personal stories of many individual actors involved in the development of AI research.
The New Fire shows how AI, and machine learning in particular, is already reshaping the way States think about war and peace. The analogy with fire is a way to suggest that AI does not have to result in the destruction of humanity and that we can learn to harness this “new fire” to ensure it used for good. The book is structured in three parts. In the first section, the three “sparks” of AI are analysed, namely data, algorithms, and computing power. Sections 2 and 3, on the other hand, focus on (inter)national security concerns, covering issues such as AI-powered cyberattacks, AI-powered disinformation campaigns, and autonomous lethal weapons.
In each chapter, emerging issues are clearly presented and assessed through three different perspectives: the pessimistic view (called the “Cassandras”), the optimistic view (the “evangelists”), and the realistic view (the “warriors”). This allows for a balanced discussion and does not prevent the authors from making their own recommendations on how democratic States, and especially the US, should go about investing their resources in AI research applied to geopolitics.
Whereas some thinkers like Yuval Noah Harari believe that AI will give autocracies an advantage over democracies, Buchanan and Imbrie are more optimistic and outline some of the strengths of democracies in today’s geopolitical competition around AI, from the control of computer chips’ manufacturing capabilities to the attraction of R&D talent. But a lot of uncertainty remains, and the many possible scenarios developed in the book are definitely helpful in grasping the potential impact of AI on the future of international relations.
I would like to thank NetGalley and MIT Press for giving me the opportunity to review this book prior to its official release in March 2022.
The New Fire shows how AI, and machine learning in particular, is already reshaping the way States think about war and peace. The analogy with fire is a way to suggest that AI does not have to result in the destruction of humanity and that we can learn to harness this “new fire” to ensure it used for good. The book is structured in three parts. In the first section, the three “sparks” of AI are analysed, namely data, algorithms, and computing power. Sections 2 and 3, on the other hand, focus on (inter)national security concerns, covering issues such as AI-powered cyberattacks, AI-powered disinformation campaigns, and autonomous lethal weapons.
In each chapter, emerging issues are clearly presented and assessed through three different perspectives: the pessimistic view (called the “Cassandras”), the optimistic view (the “evangelists”), and the realistic view (the “warriors”). This allows for a balanced discussion and does not prevent the authors from making their own recommendations on how democratic States, and especially the US, should go about investing their resources in AI research applied to geopolitics.
Whereas some thinkers like Yuval Noah Harari believe that AI will give autocracies an advantage over democracies, Buchanan and Imbrie are more optimistic and outline some of the strengths of democracies in today’s geopolitical competition around AI, from the control of computer chips’ manufacturing capabilities to the attraction of R&D talent. But a lot of uncertainty remains, and the many possible scenarios developed in the book are definitely helpful in grasping the potential impact of AI on the future of international relations.
I would like to thank NetGalley and MIT Press for giving me the opportunity to review this book prior to its official release in March 2022.
May 29, 2022
This is a very informative book for the general public. The topic is of concern to every citizen of the modern world and the authors include many interesting facts that enrich substantially one's view of the world.
I would have personally preferred the text to be more concise: the useful information in the book could have been conveyed in a smaller number of pages without it becoming too dense.
I would have personally preferred the text to be more concise: the useful information in the book could have been conveyed in a smaller number of pages without it becoming too dense.
March 16, 2022
This book is not just for the policy geeks among us but for anyone, who wants to understand how AI will shape our world and what dynamics are at play. Andrew Imbrie and Ben Buchanan do a great job in breaking down an abstract topic through stories and convincing characters. I particularly enjoyed the different schools of thought on AI that they introduced. A Must-read!
June 3, 2024
Good overview and a great place to start
November 30, 2021
A very good and timely commentary on most aspects of AI and machine learning and how it influences our daily lives. Very well researched and written. Accessible even for the novices on the topic.
July 18, 2023
This book might be of some value to people who imagine that AI is almost all hype. Such a person might see enough hints of important progress to look for more books about AI. But please don't imagine that The New Fire will be sufficient to produce a good understanding of how AI will change the world.
The book has numerous references. Alas, those references suggest that the authors got more of their ideas from the new media than from the AI literature.
The book's title is a decent metaphor for AI, in that developing AI is like playing with fire. We'll either get it to build us a better civilization, or we'll burn down the current one.
Are they talking about mushroom cloud sized fires? Or dumpster fires? My impression is that they expect something in the middle of that range. E.g. the authors imply that autonomous weapons are a more serious problem than having software that's capable of outwitting us. But at least the book warns us of the high risk that we'll end up with autonomous weapons even though we don't want them.
The authors classify people's views AI into three categories: evangelists, warriors, and Cassandras. I'm especially puzzled by the authors' description of Cassandras, who allegedly think AI will be too weak to be valuable.
I just finished writing about the XPT tournament, where viewpoints were pretty clearly divided between Cassandras who agreed that AI will be more powerful than most realize, and skeptics who think AI is a minor concern because it's mostly hype. I'm sure there are people who fit Buchanan and Imbrie's categories, but those three categories combined seem clearly in the minority.
For better introductions to AI for laymen, see Brian Christian's The Alignment Problem, and The AI Does Not Hate You, by Tom Chivers.
The book has numerous references. Alas, those references suggest that the authors got more of their ideas from the new media than from the AI literature.
The book's title is a decent metaphor for AI, in that developing AI is like playing with fire. We'll either get it to build us a better civilization, or we'll burn down the current one.
Are they talking about mushroom cloud sized fires? Or dumpster fires? My impression is that they expect something in the middle of that range. E.g. the authors imply that autonomous weapons are a more serious problem than having software that's capable of outwitting us. But at least the book warns us of the high risk that we'll end up with autonomous weapons even though we don't want them.
The authors classify people's views AI into three categories: evangelists, warriors, and Cassandras. I'm especially puzzled by the authors' description of Cassandras, who allegedly think AI will be too weak to be valuable.
I just finished writing about the XPT tournament, where viewpoints were pretty clearly divided between Cassandras who agreed that AI will be more powerful than most realize, and skeptics who think AI is a minor concern because it's mostly hype. I'm sure there are people who fit Buchanan and Imbrie's categories, but those three categories combined seem clearly in the minority.
For better introductions to AI for laymen, see Brian Christian's The Alignment Problem, and The AI Does Not Hate You, by Tom Chivers.
March 7, 2023
Great high-level overview of the next stage of geopolitical ramifications of AI's current capacities. I'm glad it went into AI's current limitations in terms of warfare - I would have loved it if the book had also gone into the developing nature of AI as a field and its own knowledge gaps (i.e. interpretability). The final chapter has a useful list of concrete policy recommendations which I expect to turn back to again.
April 25, 2022
A very engaging read on the history of AI with a lot of emphasis on deep learning, and implications for geopolitics. Must read for people interested in future conflicts and surveillance. The authors provided nuanced insights on where AI systems may fail that we often don’t hear about in mainstream circles.
November 9, 2022
Extremely accessible read for non-technical folks. It starts by introducing types of machine learning systems and technical concepts to give the reader a baseline understanding before exploring specific applications. There's a good mix of technical explanations, historical context, and anecdotes that make it a rather easy read, and features a heavy focus on AI competition with China.
July 20, 2022
In depth knowledge about AI and how everything started and how it could be used for destruction and creation.
July 26, 2022
Pretty technical and thorough.
February 19, 2023
No es súper entretenido ni llevadero, aunque tiene algo de info que está buena.
October 1, 2023
The New Fire: War, Peace, and Democracy in the Age of AI by Ben Buchanan and Andrew Imbre is a very good primer. Highly recommended for people interested in cyber security and AI research.
September 11, 2023
This is a worthy addition to the recent plethora of books on AI.
Data, algorithms and computing power are the three sparks that have created this new fire.
This book explores in particular the militarized use of AI - autonomous weapons, smart defence systems and cyberwarfare.
The potential for these to be abused by autocratic regimes is a serious threat to democracies who are bound by ethical constraints.
* Re-reading this in the ChatGPT world of 2023 provided a different and more urgent perspective to the information contained in this book.
3 characters are used: evangelists, warriors and Cassandras - those that want to harness AI for the good of all, those that want to use it for national security and those that fear it.
All 3 voices are important in light of AI now hitting exponential adoption.
Data, algorithms and computing power are the three sparks that have created this new fire.
This book explores in particular the militarized use of AI - autonomous weapons, smart defence systems and cyberwarfare.
The potential for these to be abused by autocratic regimes is a serious threat to democracies who are bound by ethical constraints.
* Re-reading this in the ChatGPT world of 2023 provided a different and more urgent perspective to the information contained in this book.
3 characters are used: evangelists, warriors and Cassandras - those that want to harness AI for the good of all, those that want to use it for national security and those that fear it.
All 3 voices are important in light of AI now hitting exponential adoption.
Displaying 1 - 15 of 15 reviews