What do you think?
Rate this book
The Handover: How We Gave Control of Our Lives to Corporations, States and AIs
Countless books, news reports, and opinion pieces have announced the impending arrival of artificial intelligence, with most claiming that it will upend our world, revolutionizing not just work but society overall. Yet according to political philosopher and historian David Runciman, we’ve actually been living with a version of AI for 300 years, because states and corporations are robots, too. In The Handover, Runciman explains our current situation through the history of these “artificial agents” we created to rescue us from our all-too-human limitations—and demonstrates what this radical new view of our recent past means for our collective future.
From the United States and the United Kingdom to the East India Company, Standard Oil, Facebook, and Alibaba, states and corporations have gradually, and then much more rapidly, taken over the planet. They have helped to conquer poverty and eliminate disease, but also unleashed global wars and environmental degradation. As Runciman demonstrates, states and corporations are the ultimate decision-making machines, defined by their ability to make their own choices and, crucially, to sustain the consequences of what has been chosen. And if the rapid spread of the modern state and corporation has already transformed the conditions of human existence, new AI technology promises the same.
But what happens when AI interacts with other kinds of artificial agents, the inhuman kind represented by states and corporations? Runciman argues that the twenty-first century will be defined by increasingly intense battles between state and corporate power for the fruits of the AI revolution. In the end, it is not our own, human relationship with AI that will determine our future. Rather, humanity’s fate will be shaped by the interactions among states, corporations, and thinking machines.
With clarity and verve, The Handover presents a brilliantly original history of the last three centuries and a new understanding of the immense challenges we now face.
From the United States and the United Kingdom to the East India Company, Standard Oil, Facebook, and Alibaba, states and corporations have gradually, and then much more rapidly, taken over the planet. They have helped to conquer poverty and eliminate disease, but also unleashed global wars and environmental degradation. As Runciman demonstrates, states and corporations are the ultimate decision-making machines, defined by their ability to make their own choices and, crucially, to sustain the consequences of what has been chosen. And if the rapid spread of the modern state and corporation has already transformed the conditions of human existence, new AI technology promises the same.
But what happens when AI interacts with other kinds of artificial agents, the inhuman kind represented by states and corporations? Runciman argues that the twenty-first century will be defined by increasingly intense battles between state and corporate power for the fruits of the AI revolution. In the end, it is not our own, human relationship with AI that will determine our future. Rather, humanity’s fate will be shaped by the interactions among states, corporations, and thinking machines.
With clarity and verve, The Handover presents a brilliantly original history of the last three centuries and a new understanding of the immense challenges we now face.
352 pages, Hardcover
First published September 7, 2023
38 people are currently reading
663 people want to read
About the author
David Runciman
25 books174 followersDavid Runciman teaches politics at Cambridge.
He writes regularly about politics and current affairs for a wide range of publications including the London Review of Books. The author of several books, he also hosted the widely-acclaimed podcast Talking Politics, along with the series ‘History of Ideas’. Past Present Future* is his new weekly podcast, where he is exploring the history of ideas from politics to philosophy, culture to technology.
*Ideas from the past, questions about the present, shaping the future.
He writes regularly about politics and current affairs for a wide range of publications including the London Review of Books. The author of several books, he also hosted the widely-acclaimed podcast Talking Politics, along with the series ‘History of Ideas’. Past Present Future* is his new weekly podcast, where he is exploring the history of ideas from politics to philosophy, culture to technology.
*Ideas from the past, questions about the present, shaping the future.
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 28 of 28 reviews
August 21, 2023
Fascinating and less dark than you would imagine given the subtitle. Raises some of those issues that should have been obvious but yet may not have occurred to you before about how artificial, essentially imaginary entities like states and corporations long outlast the people who created them and now essentially control the world. Manages to study their impact through a reasonable, non-conspiratorial worldview, particularly impressive once he gets into AI towards the end. Also appreciate Runciman's analysis of how ancient empires differed from their modern equivalents mostly due to time and distance. Empires like the Roman were "an imagined community held together at its edges by nothing more than distant rumours of what it all meant at the centre."
January 23, 2025
No I would not like to harness the power of AI to streamline my workload thank you very much.
October 29, 2023
The idea explored in this book is that states and corporations are machines. Runciman argues that the reason for the large scale technological developments over the past 300 years is because of the widespread implementation of these machines: states provide a stable environment and funding for scientific research to take place, and corporations hone the processes designed by scientists to making them more efficient with the aim of making profit. These machines are artificial in the sense that they are easily replicable and are autonomous agents as they do not depend on certain individuals, and parallels are drawn with recent developments in AI. The argument is that these properties of modern states and corporations separates them from previous modes of government and economic setups in, say, the Roman empire. The idea is reminiscent of the Gaia hypothesis (Novacene) where the whole world is viewed as one big machine. Runciman is essentially describing how this machine is split into submachines.
The problem I have with this book is that I don't think Runciman goes far enough. He argues that states and corporations are not superhuman or superintelligent in the sense that people are concerned about with regards to AI taking over the world. His argument is essentially that they are run by people and so by definition they can only be as intelligent as the people running them. I disagree with this point of view. The idea that a machine is only as clever as its component parts doesn't make sense. Humans are more intelligent than their neurons and ChatGPT is more intelligent than its transistors. Runciman also argues that current AI like ChatGPT cannot be intelligent on the basis that it only tries to come up with the most probable answer to a prompt. While I agree with this to an extent, it's not at all obvious that a machine that's designed in this way can't be intelligent. We can't look at the design of a machine to decide whether or not it is intelligent.
The only approach that makes sense to determine if something is superintelligent is to use a sort of Turing test. Of course, we can't interact directly with states and corporations and so the Turing test for these artificial agents should be whether or not they behave in the way that AI researchers are concerned that artificial general intelligence might behave, see for example Human Compatible. These are things like alignment issues where the machines start to behave in a way that the designer didn't intend or the AI systems manipulating their human controllers for their own gain. By these metrics states and corporations are undoubtedly superintelligent. We can't turn off the United States. Humans die in conflicts caused by disagreements between states. The existence of nuclear weapons poses an existential threat to the planet which humans cannot control. Runciman draws parallels between the technological and economic revolution of the past 300 years and the potential for a second revolution with the advent of thinking machines. I think it's a mistake to separate these revolutions. Thinking machines can exist only because they have been created by states and corporations.
If you accept that states and corporations are the AGI systems, then you can combine thinking from history and AI saftey. It might be possible to change the design states and corporations in the ways suggested by AI researchers to prevent humans losing control of them, and since these AGI systems haven't killed us yet, AI researchers could use states and corporations as models for how AI systems might develop. This is what I wanted the book to be about.
The problem I have with this book is that I don't think Runciman goes far enough. He argues that states and corporations are not superhuman or superintelligent in the sense that people are concerned about with regards to AI taking over the world. His argument is essentially that they are run by people and so by definition they can only be as intelligent as the people running them. I disagree with this point of view. The idea that a machine is only as clever as its component parts doesn't make sense. Humans are more intelligent than their neurons and ChatGPT is more intelligent than its transistors. Runciman also argues that current AI like ChatGPT cannot be intelligent on the basis that it only tries to come up with the most probable answer to a prompt. While I agree with this to an extent, it's not at all obvious that a machine that's designed in this way can't be intelligent. We can't look at the design of a machine to decide whether or not it is intelligent.
The only approach that makes sense to determine if something is superintelligent is to use a sort of Turing test. Of course, we can't interact directly with states and corporations and so the Turing test for these artificial agents should be whether or not they behave in the way that AI researchers are concerned that artificial general intelligence might behave, see for example Human Compatible. These are things like alignment issues where the machines start to behave in a way that the designer didn't intend or the AI systems manipulating their human controllers for their own gain. By these metrics states and corporations are undoubtedly superintelligent. We can't turn off the United States. Humans die in conflicts caused by disagreements between states. The existence of nuclear weapons poses an existential threat to the planet which humans cannot control. Runciman draws parallels between the technological and economic revolution of the past 300 years and the potential for a second revolution with the advent of thinking machines. I think it's a mistake to separate these revolutions. Thinking machines can exist only because they have been created by states and corporations.
If you accept that states and corporations are the AGI systems, then you can combine thinking from history and AI saftey. It might be possible to change the design states and corporations in the ways suggested by AI researchers to prevent humans losing control of them, and since these AGI systems haven't killed us yet, AI researchers could use states and corporations as models for how AI systems might develop. This is what I wanted the book to be about.
May 29, 2024
I really like Runciman’s style of writing and there are some really revelatory parts of this book, but overall the main points are pretty well defined early on and other than the odd snippet not much was added in the later parts.
That said it’s a pretty under-explored idea - the similarity between states, corporations and computers - and proves a useful lens here.
That said it’s a pretty under-explored idea - the similarity between states, corporations and computers - and proves a useful lens here.
November 28, 2025
????????????????? read this to feel educated but now I just feel dumb.
December 10, 2023
Originalus požiūris į tai, kas vyskta su pasauliniais procesais. Valstybės, korporacijos, žmonės grupėse - visa tai jau yra mechanizmai gyvenantys atskirtus gyvenimus ir turintys savo mechanizmus. Todėl ateinantis dirbtinis intelektas nėra kažkas visiškai nauja.
Knygos problema- labai akademiškas dėstymas.
Knygos problema- labai akademiškas dėstymas.
September 11, 2024
In 'The Handover,' David Runciman tackles one of the most pressing issues of our time: the gradual transfer of control from individuals to larger entities like governments, corporations, and artificial intelligence. This thought-provoking work examines the delicate balance between collective power and individual autonomy in our rapidly evolving world.
Key Themes:
The Modern Leviathan: Runciman explores how the metaphorical "Leviathan" of government power has shaped modern society, creating tension between group decisions and individual choices.
Corporate Dominance: The book offers a striking analysis of how corporations now wield influence that often rivals or surpasses traditional government authority.
Individual vs. Collective: The author challenges readers to consider the trade-offs between collective decision-making (which can provide security and solve large-scale problems) and the preservation of personal autonomy.
The AI Question: Runciman posits that we're at a crucial juncture in history, facing critical decisions about how much control to cede to artificial intelligence—choices that could determine humanity's fate.
Runciman's intelligent writing style forces readers to reconsider what it means to live under systems that prioritize collective stability over individual agency. He examines these complex issues from multiple angles, careful not to advocate for any single solution. This book is both intellectually stimulating and deeply unsettling. Runciman challenges us to think critically about our role in shaping—or surrendering—our future. As we navigate an increasingly interconnected and automated world, "The Handover" provides a crucial framework for understanding the power dynamics that influence our lives.
'The Handover' is particularly relevant for those interested in political philosophy, corporate influence on society, the future of AI governance, and the balance between personal freedom and collective security. Whether you're a policy maker, a tech enthusiast, or simply a concerned citizen, this book offers valuable insights into the systems that govern our world and the choices we face as a society.
Key Themes:
The Modern Leviathan: Runciman explores how the metaphorical "Leviathan" of government power has shaped modern society, creating tension between group decisions and individual choices.
Corporate Dominance: The book offers a striking analysis of how corporations now wield influence that often rivals or surpasses traditional government authority.
Individual vs. Collective: The author challenges readers to consider the trade-offs between collective decision-making (which can provide security and solve large-scale problems) and the preservation of personal autonomy.
The AI Question: Runciman posits that we're at a crucial juncture in history, facing critical decisions about how much control to cede to artificial intelligence—choices that could determine humanity's fate.
Runciman's intelligent writing style forces readers to reconsider what it means to live under systems that prioritize collective stability over individual agency. He examines these complex issues from multiple angles, careful not to advocate for any single solution. This book is both intellectually stimulating and deeply unsettling. Runciman challenges us to think critically about our role in shaping—or surrendering—our future. As we navigate an increasingly interconnected and automated world, "The Handover" provides a crucial framework for understanding the power dynamics that influence our lives.
'The Handover' is particularly relevant for those interested in political philosophy, corporate influence on society, the future of AI governance, and the balance between personal freedom and collective security. Whether you're a policy maker, a tech enthusiast, or simply a concerned citizen, this book offers valuable insights into the systems that govern our world and the choices we face as a society.
December 13, 2023
The book gives us a difficult read. The second label of the book makes a small explanation of what Runciman is doing. Many of the pages drive us to thinking about what might be coming in the world. It’s a book that is best understood when you are in a small group of people who are reading this book. And there’s the need to get the book when we can talk in a group of people who are reading it. On page 17, we get “The relationship between human beings, the groups they form and the machines they build is at the heart of this book.” (p. 17)
In Chapter 2, “Groupthink,” “We move from consciousness to unconsciousness in the creation of group minds…” It’s a challenge. But this chapter uses the Twelve Angry Men as a example of the theme.
He also uses the differences between states and corporations as the major changes in our world. He also tells us about the events of the 1800’s that started a change that continues in how we live and think. But there are some references to what happened in the Roman Empire. And then there’s data that the South Koreans had a life like the live of England in the year 1300. (p. 136) He also shows how the growth of America ws done by Black people. (p. 163)
Yes, it’s a difficult book to read. I should have read it with a person who had my interests. But, it the reading was powerful.
In Chapter 2, “Groupthink,” “We move from consciousness to unconsciousness in the creation of group minds…” It’s a challenge. But this chapter uses the Twelve Angry Men as a example of the theme.
He also uses the differences between states and corporations as the major changes in our world. He also tells us about the events of the 1800’s that started a change that continues in how we live and think. But there are some references to what happened in the Roman Empire. And then there’s data that the South Koreans had a life like the live of England in the year 1300. (p. 136) He also shows how the growth of America ws done by Black people. (p. 163)
Yes, it’s a difficult book to read. I should have read it with a person who had my interests. But, it the reading was powerful.
December 15, 2025
As the subtitle suggests, Professor David Runciman is putting AIs on an equal footing with states and corporations -- they are all "machines," or "automata," as Thomas Hobbes would say. Therefore, a lot of existing ideas in political philosophy (Karl Marx and Hannah Arendt appear a lot, in addition to Hobbes) can be put to use in thinking about AIs. When I started this book, I read it as a great framework for exploring humanity's relationship with AIs. Now as I reflect on the central message, and noting much of the book is also dedicated to how the machines before AIs work, I can see it is a cautionary tale of how we as humans could give up, or are already giving up our agency, purpose, and wellbeing -- handing them over to the machines. I enjoyed this book, and learned many insights from the surprising amount of details about how all these types of machines are similar and different.
There are a few ways to approach the central message of the book. One is to bring out the title of the last chapter: "Who Works for Whom?" Clearly, the author intends to warn us against having humans work exclusively for the perpetuation of machines, which has already been happening considering the perverted incentives in empty nationalism and where businesses justify maximizing corporate profits. Another is my favorite quote, from Chapter 2: "are we looking to them for an answer, or are we looking to them for a decision?" When we confuse answers with decisions, we forget that machines are working for us, as tools to provide us with answers, to help us make decisions; we risk letting the machines make decisions for us, and we start to work for them. Yet another quote from Chapter 5 also works: "The modern age has always been torn between our twin impulses towards security and emancipation." People seeking security will voluntarily give up their agency and freedom, and let the machines decide for them -- an age-old dilemma that always been explored in literature and philosophy (e.g. The Grand Inquisitor). They all lead to the same concern contained in the title of this book.
There are some notable ideas contributing to the central message that are themselves interesting. One is: "Any question can invite an answer, or it can invite a decision." If we confuse what we're asking, we confuse answers with decisions; and we interpret and act on the machines' responses wrongly. An inspiring application of this is to democracy: states work with decisions, not answers:
A related idea is: "intelligence is not decision-making, just as information is not a choice." We can ask the machines for answers, and they will give us intelligence and information. But it is us who should make decisions and choices. This reminds me of a Chinese idiom: 尽信书不如无书, or "believing everything in books is worse than having no books at all." All of these are cautioning not to blindly make the decision of "following any answer you get."
Another important idea is to realize the different levels of agency among different machines. States have way more agency than small corporations. Large international corporations start to have more and more agency. These are a result of "humans have found it very helpful to let them decide certain things for us," but they'll start to have lives of their own: "Without our input they will make their own choices. ... It is fanciful to suppose that they would choose to phase us out once we did not serve their purposes any more." How so? The author gave an impressive example:
There is some nuance to this, since it doesn't mention here the people that benefited from the machines. We can say that humanity changed our way of living to adapt to the machines, but they in turn do benefit many of us. So it's not really fair to say humans are "sacrificed" for the machine -- maybe more like they are also "sacrificed" for the benefit of some other people. What's important is that the people who worked for the machines and the people that the machines worked for are not necessarily the same group; for different machines and at different times, one person can be on different sides of the working-for relationships. It is a very intertwined issue and it's very hard to separate the people from the machines if we only talk about "humans vs machines" collectively instead of case by case. I think the author doesn't do a good job in this regard.
The author does discuss how the machines are different from the components, often the people. This distinction has moral consequences, since when interpreting and assigning agency and accountability, it is not straightforward whether it should be to the machine or to the people. Regarding this, there are existing moral philosophical discussions in situations of collective decision-making and crowd intelligence. These are topics that the author touches upon, but unfortunately not in any substantive detail. Often, the book's discussions risk claiming a false dichotomy, that there should be only 1 center of culpability or moral responsibility. Since in many cases, both the individuals and the group or the machine are needed to achieve the results, we could have held all parties liable: the individuals as well as the collective. There isn't any mention of whether it is feasible to set up the rules and processes for both sides to be accountable.
Regarding group identity, the author makes a clear distinction between AIs as machines and states and corporations as machines. He points out that AIs as machines do not consists of humans, so any discussions about group responsibility does not apply. I think this is really underexplored until towards the end of the book, where the author aggregates the machines together: states, corporations, and AIs working together collectively to achieve societal impact. I don't think it's fair for the author to make this distinction, since AIs have always been working together with at least some form of human organizations for it to be able to be deployed. Therefore when in general people talk about the societal impact of AIs, such as in most cases in this book, they're really talking about the collective AI-corporation complexes. At the same time, people also call the front-end interfaces the AIs and forget the corporate entities working behind that make them functional; when they discuss the technological details and the neutrality of technology, for example, they confuse AIs with the AI-corporation complexes. The discussions about group responsibility does apply to AIs for the purpose of this book: it applies to the AI-corporation complexes. In the AI-corporation complexes, the role of AIs is a similar but more powerful extension of the websites or the legal documents of the existing machines: states and corporations -- AIs are executing the agencies of the human organizations. This is true for AIs today, as so far there has been no clear evidence that AIs can have agencies of their own; but if the author wants to warn about future AIs, that can be different.
Another distinction for AIs as machines I see the author implicitly makes is that he discusses whether the robots, or AIs, have rights; at the same time, he takes the rights of states and corporations as given, which admittedly is a historical fact. Given the amount of history the author has dived into, it is somewhat surprising that he argues that robots don't have rights because they can't feel, yet he does not question that states and corporations have rights in the same way. Perhaps the answer to any of these should depend on the usefulness, which the author mentions in various places in the book but fails to make the connection.
Other insights I like in the book include the distinctions among career, jobs, and tasks (or work vs. labour, as Hannah Arendt would put it):
The author also notes that there are other machines human have made in addition to the main ones he is writing about. Presumably, the Church is a machine, too, and it should've played a very important role that I feel is missing from the current picture.
There are a few ways to approach the central message of the book. One is to bring out the title of the last chapter: "Who Works for Whom?" Clearly, the author intends to warn us against having humans work exclusively for the perpetuation of machines, which has already been happening considering the perverted incentives in empty nationalism and where businesses justify maximizing corporate profits. Another is my favorite quote, from Chapter 2: "are we looking to them for an answer, or are we looking to them for a decision?" When we confuse answers with decisions, we forget that machines are working for us, as tools to provide us with answers, to help us make decisions; we risk letting the machines make decisions for us, and we start to work for them. Yet another quote from Chapter 5 also works: "The modern age has always been torn between our twin impulses towards security and emancipation." People seeking security will voluntarily give up their agency and freedom, and let the machines decide for them -- an age-old dilemma that always been explored in literature and philosophy (e.g. The Grand Inquisitor). They all lead to the same concern contained in the title of this book.
There are some notable ideas contributing to the central message that are themselves interesting. One is: "Any question can invite an answer, or it can invite a decision." If we confuse what we're asking, we confuse answers with decisions; and we interpret and act on the machines' responses wrongly. An inspiring application of this is to democracy: states work with decisions, not answers:
"States are constituted by how they arrive at decisions, not by whether those decisions are good or bad ones. If the wrong decisions keep getting made, that means changing the how. Simply coming up with better solutions won’t make the difference. Yet changing the how – alternative voting systems, bureaucratic reform, deliberative assemblies, constitutional design – is pretty unsexy. People who want to do good rarely work on these issues.
...
Still, would you rather have the idea, or be one of the regulators whose job it is to make sure it doesn’t get abused? Would you rather be speculating about the what, or tinkering with the how? The danger is that the smart people end up doing the former, but the latter still determines where we end up."
A related idea is: "intelligence is not decision-making, just as information is not a choice." We can ask the machines for answers, and they will give us intelligence and information. But it is us who should make decisions and choices. This reminds me of a Chinese idiom: 尽信书不如无书, or "believing everything in books is worse than having no books at all." All of these are cautioning not to blindly make the decision of "following any answer you get."
Another important idea is to realize the different levels of agency among different machines. States have way more agency than small corporations. Large international corporations start to have more and more agency. These are a result of "humans have found it very helpful to let them decide certain things for us," but they'll start to have lives of their own: "Without our input they will make their own choices. ... It is fanciful to suppose that they would choose to phase us out once we did not serve their purposes any more." How so? The author gave an impressive example:
"cars lack plenty of kinds of adaptability compared with horses: they can’t step over obstacles, or move sideways, or swim through streams. That didn’t stop us building an entire economy around them, and road networks suited to all their limitations, sacrificing many hundreds of thousands of our own lives in the process."
There is some nuance to this, since it doesn't mention here the people that benefited from the machines. We can say that humanity changed our way of living to adapt to the machines, but they in turn do benefit many of us. So it's not really fair to say humans are "sacrificed" for the machine -- maybe more like they are also "sacrificed" for the benefit of some other people. What's important is that the people who worked for the machines and the people that the machines worked for are not necessarily the same group; for different machines and at different times, one person can be on different sides of the working-for relationships. It is a very intertwined issue and it's very hard to separate the people from the machines if we only talk about "humans vs machines" collectively instead of case by case. I think the author doesn't do a good job in this regard.
The author does discuss how the machines are different from the components, often the people. This distinction has moral consequences, since when interpreting and assigning agency and accountability, it is not straightforward whether it should be to the machine or to the people. Regarding this, there are existing moral philosophical discussions in situations of collective decision-making and crowd intelligence. These are topics that the author touches upon, but unfortunately not in any substantive detail. Often, the book's discussions risk claiming a false dichotomy, that there should be only 1 center of culpability or moral responsibility. Since in many cases, both the individuals and the group or the machine are needed to achieve the results, we could have held all parties liable: the individuals as well as the collective. There isn't any mention of whether it is feasible to set up the rules and processes for both sides to be accountable.
Regarding group identity, the author makes a clear distinction between AIs as machines and states and corporations as machines. He points out that AIs as machines do not consists of humans, so any discussions about group responsibility does not apply. I think this is really underexplored until towards the end of the book, where the author aggregates the machines together: states, corporations, and AIs working together collectively to achieve societal impact. I don't think it's fair for the author to make this distinction, since AIs have always been working together with at least some form of human organizations for it to be able to be deployed. Therefore when in general people talk about the societal impact of AIs, such as in most cases in this book, they're really talking about the collective AI-corporation complexes. At the same time, people also call the front-end interfaces the AIs and forget the corporate entities working behind that make them functional; when they discuss the technological details and the neutrality of technology, for example, they confuse AIs with the AI-corporation complexes. The discussions about group responsibility does apply to AIs for the purpose of this book: it applies to the AI-corporation complexes. In the AI-corporation complexes, the role of AIs is a similar but more powerful extension of the websites or the legal documents of the existing machines: states and corporations -- AIs are executing the agencies of the human organizations. This is true for AIs today, as so far there has been no clear evidence that AIs can have agencies of their own; but if the author wants to warn about future AIs, that can be different.
Another distinction for AIs as machines I see the author implicitly makes is that he discusses whether the robots, or AIs, have rights; at the same time, he takes the rights of states and corporations as given, which admittedly is a historical fact. Given the amount of history the author has dived into, it is somewhat surprising that he argues that robots don't have rights because they can't feel, yet he does not question that states and corporations have rights in the same way. Perhaps the answer to any of these should depend on the usefulness, which the author mentions in various places in the book but fails to make the connection.
Other insights I like in the book include the distinctions among career, jobs, and tasks (or work vs. labour, as Hannah Arendt would put it):
"Performing tasks is what machines are good at. The better they get at it, the more work becomes task orientated. ...
At the same time, smart machines are helping to change the character of the states and corporations that provide the jobs for people. Just as technology can fragment the personal identity of individuals into a series of data points, so it can fragment the personal identity of states and corporations into a series of tasks, or projects. Give a machine work and it will become a task; give a human work and it can become a job. Inevitably, the more work that machines do, the less the identity of artificial persons will be shaped by the secure jobs they are able to sustain, because those jobs will have been replaced by tasks. Instead, even states will be identified by the different projects they undertake, each with its own discrete timeframe and end point.
As careers get reduced to jobs, likewise jobs get reduced to tasks. More and more people are employed on short-term contracts, offered by organisations that see work as something to be understood in terms of its outcome rather than as an ongoing relationship. This is only in part because people are being replaced by machines. It is also because people are increasingly being employed like machines."
The author also notes that there are other machines human have made in addition to the main ones he is writing about. Presumably, the Church is a machine, too, and it should've played a very important role that I feel is missing from the current picture.
May 18, 2024
"The state is an artificial general agent. Its remit is relatively limitless, even if its intelligence is highly limited. What it might do – what it might want – what it might need – remain key questions both for our future and for the future of the technology that has the power to reshape our lives. Humans make states, which can acquire a life of their own; states make corporations, which can acquire a life of their own; corporations make robots, which do not have a life of their own yet. What comes next depends on what states and corporations might do with the powers they have been given, and what we might still be able to do to shape how they use them."
"Let’s not kid ourselves that the age of AI – with the looming challenges of black-box decision-making and algorithmic procedures whose outcomes are a mystery even to their creators – poses a unique challenge for human understanding. Just look at your own life. Do you fully understand where the group decisions come from that shape who you are? Do you understand even when those decisions are ones you have been part of yourself? I thought not. The fact that groups are made out of people doesn’t make them any easier to see inside than other kinds of machines. The black boxes are all around us already. In one sense, at least, they are us."
"States and corporations reflect two different sides of our contemporary fear of machines that have escaped human control. One is that we will build machines that we don’t know how to switch off, either because we have become too dependent on them or because we can’t find the off switch. That’s states. The other is that we build machines that self-replicate in ways that we can no longer regulate. They start spewing out versions of themselves to the point where we are swamped by them. That’s corporations."
"Yet we are still promised that the truly big transformation is at hand – a time is coming soon when the human experience will be upended again by new machines whose superhuman qualities will massively enhance what we can achieve. These intelligent machines will transform life expectancy – from decades to centuries – and travel – from international to interplanetary – and human memory – from limited to limitless. We won’t recognise ourselves."
"Let’s not kid ourselves that the age of AI – with the looming challenges of black-box decision-making and algorithmic procedures whose outcomes are a mystery even to their creators – poses a unique challenge for human understanding. Just look at your own life. Do you fully understand where the group decisions come from that shape who you are? Do you understand even when those decisions are ones you have been part of yourself? I thought not. The fact that groups are made out of people doesn’t make them any easier to see inside than other kinds of machines. The black boxes are all around us already. In one sense, at least, they are us."
"States and corporations reflect two different sides of our contemporary fear of machines that have escaped human control. One is that we will build machines that we don’t know how to switch off, either because we have become too dependent on them or because we can’t find the off switch. That’s states. The other is that we build machines that self-replicate in ways that we can no longer regulate. They start spewing out versions of themselves to the point where we are swamped by them. That’s corporations."
"Yet we are still promised that the truly big transformation is at hand – a time is coming soon when the human experience will be upended again by new machines whose superhuman qualities will massively enhance what we can achieve. These intelligent machines will transform life expectancy – from decades to centuries – and travel – from international to interplanetary – and human memory – from limited to limitless. We won’t recognise ourselves."
November 21, 2023
Maddening at times and very enlightening in others.
David’s key thesis is that we already live in a world run by inhuman machines that are very comfortable interacting with each other: States and corporations. Given this how should we think about the possibility of a new machine - super powerful AI - entering this two-way fray?
Sometimes he forces this vision of the world a bit too strongly, others it really works for me. Full of really good thinking, but his stuff always is.
David’s key thesis is that we already live in a world run by inhuman machines that are very comfortable interacting with each other: States and corporations. Given this how should we think about the possibility of a new machine - super powerful AI - entering this two-way fray?
Sometimes he forces this vision of the world a bit too strongly, others it really works for me. Full of really good thinking, but his stuff always is.
January 11, 2024
This fascinating and original book discusses how throughout history artificial constructs such as states, corporations and now artificial intelligence help the world become better but could be a threat if not kept under control. In today’s world threats such as nuclear warfare, climate change or biological disasters are more devastating but potentially killer robots should not be ignored. With plenty of historical precedents it stresses the need for human control.
January 19, 2025
One of the most surprising books I've read in a long time. I thought more of the focus would be on the implementations of AI and how it's changing governments and corporations, but instead the author shifted the focus of AI to just another step in a system we are already involved in, non-human entities like states and corporations. The interaction is not simply us and AI, but us and corporations and states and AI and how they will all interact.
June 23, 2025
This is not an easy read. You have to grapple with dense technical concepts. It is not a one time read, you really need to read it slowly, discuss then reread. It offers some material worth examining on states, corporations, and AI. It is definitely not for everybody and perhaps that is its biggest flaw. Many will start then abandon it. It is just not going to be widely read, That is too bad because the ideas and decisions need to be widely considered rather than in the hands of the few.
January 21, 2024
Presented some interesting ideas about the state as person vs as machine, and how the rise of AI (not yet AGI) leaves us with a choice to make on the nature of our future governance and systems. Perhaps as a symptom of listening to this one rather than reading it, I found the ideas interesting but the overall thread of the argument difficult to follow at times.
February 24, 2024
Despite being marketed as such, this book is not primarily about AI (at least in the technical sense) and its takeover of our societies. Rather, it is a narrative about the history of artificial agency in the sphere of social and political organisation, where Runciman’s focus is on states and corporations.
Ruciman’s point of departure, as per usual, is Hobbes’s Leviathan. It is this analysis of the creation of an artificial man in the middle of the English Civil War, - an algorithm for collective decision-making - which produced the longest-lasting artificial agency in history - a modern bureaucratic state.
Runciman expends a great amount of space explaining this dichotomy between ‘decisions’ and ‘answers’ which is important for his latter argument about the joining up of human intelligence, artificial intelligence and what he calls the ‘artificial agency’ (states and corporations).
The core argument of the book is that we have created the artificial agency of the state, as a permanent being (or rather machine) with a universal and general field of operation - which is distinct from most corporations and historical forms of collective organisation that were created for a specific purpose or have more limited timespan, even if they are older than states they operate within (they are essentially project-based). It is the “grinding, self-sustaining qualities of established institutions” which allow for a sense of unbroken continuity, which in turn allows for human and economic progress.
The state can and most likely will outlive its citizens, and the humans that constitute it, and can take obligations based on its ability to control violence and levy taxes. By contrast, corporations, even if they are centuries old, do not have the permanence baked into the reality of their existence - but their response to different kinds of incentives can work in conjunction with the states (bureaucratic machines can talk to other kinds of bureaucratic machines) and produce technological progress.
This provides a very interesting conceptualisation of some of the most frustrating parts of the modern state, which Runciman greatly reflects on in the latter chapters. Wanting the state to function like a startup or an agile enterprise, is almost as naive as wanting it to be empathetic or personal. It can seem to be, through its temporary leaders, who may use this in their political work - but it will always be merely a charade. For Runciman, as in all of his works, the good politicians see themselves as “representatives of an impersonal state”, whereas the bad ones believe they can humanise the levers of government by subordinating them to their designs.
On the other hand, this analysis demonstrates a different perspective on Runciman’s thinking about the role and functioning of the state. In his discussion on the role of the modern state, imperialism is largely glossed over, as is the understanding of climate change as anything beyond a result of ‘wrong answers’ by impersonal corporate agents, working within badly pre-defined rules. This allows his narrative of the modern state as a largely benign force, that allows for calmer and longer-term decision-making and processes, largely devoid of the madness of the crowds leading to social uphevals (although he acknowledges possibilities for elite groupthink).
There are many further limitations of his argument, beyond the scope of this review - particularly when looking from a perspective of smaller states that, for instance also function within supranational entities such as the European Union (which he calls a mere ‘legal order’). While a large part of his analysis holds across many states, it is limited in seeing the state from the minimalist liberal tradition.
His argument ultimately rests on the understanding of the necessity for the existence of the state through the Hobbesian lenses - as otherwise life would be 'solitary, poor, nasty, brutish, and short'. But one does not need to be an anarchist to see, that both anthropological and archeological evidence shows that this was not and is not ultimately true. State as an arbiter and enforcer or structural violence is a historical choice - even if one can understand and empathise with a great deal of Runciman's analysis.
The final chapters deal with the rise of AI and this part of the discussion is quite novel, as it incorporates the artificiality of the functioning of AIs within his narrative about the first singularity occurring during the rise of the modern state. But even here he operates with terms like ‘better answers’ that are ultimately very normative, even if he does not provide full answers or prescriptions about the better organisation of society. Rather, he considers what various mixes of Human Intelligence (HI) – Artificial Intelligence (AI) – and Artificial Agency (AA) mean.
While the book may be billed as a discourse on the future of AI, its primary objective is to uphold the enduring principles of an elite-centred variation of liberal ideology. It advocates for embracing the inherent detachment from centres of authority, reassuring individuals that such detachment serves their best interests. That is not to say that I did not enjoy reading it.
Ruciman’s point of departure, as per usual, is Hobbes’s Leviathan. It is this analysis of the creation of an artificial man in the middle of the English Civil War, - an algorithm for collective decision-making - which produced the longest-lasting artificial agency in history - a modern bureaucratic state.
Runciman expends a great amount of space explaining this dichotomy between ‘decisions’ and ‘answers’ which is important for his latter argument about the joining up of human intelligence, artificial intelligence and what he calls the ‘artificial agency’ (states and corporations).
The core argument of the book is that we have created the artificial agency of the state, as a permanent being (or rather machine) with a universal and general field of operation - which is distinct from most corporations and historical forms of collective organisation that were created for a specific purpose or have more limited timespan, even if they are older than states they operate within (they are essentially project-based). It is the “grinding, self-sustaining qualities of established institutions” which allow for a sense of unbroken continuity, which in turn allows for human and economic progress.
The state can and most likely will outlive its citizens, and the humans that constitute it, and can take obligations based on its ability to control violence and levy taxes. By contrast, corporations, even if they are centuries old, do not have the permanence baked into the reality of their existence - but their response to different kinds of incentives can work in conjunction with the states (bureaucratic machines can talk to other kinds of bureaucratic machines) and produce technological progress.
This provides a very interesting conceptualisation of some of the most frustrating parts of the modern state, which Runciman greatly reflects on in the latter chapters. Wanting the state to function like a startup or an agile enterprise, is almost as naive as wanting it to be empathetic or personal. It can seem to be, through its temporary leaders, who may use this in their political work - but it will always be merely a charade. For Runciman, as in all of his works, the good politicians see themselves as “representatives of an impersonal state”, whereas the bad ones believe they can humanise the levers of government by subordinating them to their designs.
On the other hand, this analysis demonstrates a different perspective on Runciman’s thinking about the role and functioning of the state. In his discussion on the role of the modern state, imperialism is largely glossed over, as is the understanding of climate change as anything beyond a result of ‘wrong answers’ by impersonal corporate agents, working within badly pre-defined rules. This allows his narrative of the modern state as a largely benign force, that allows for calmer and longer-term decision-making and processes, largely devoid of the madness of the crowds leading to social uphevals (although he acknowledges possibilities for elite groupthink).
There are many further limitations of his argument, beyond the scope of this review - particularly when looking from a perspective of smaller states that, for instance also function within supranational entities such as the European Union (which he calls a mere ‘legal order’). While a large part of his analysis holds across many states, it is limited in seeing the state from the minimalist liberal tradition.
His argument ultimately rests on the understanding of the necessity for the existence of the state through the Hobbesian lenses - as otherwise life would be 'solitary, poor, nasty, brutish, and short'. But one does not need to be an anarchist to see, that both anthropological and archeological evidence shows that this was not and is not ultimately true. State as an arbiter and enforcer or structural violence is a historical choice - even if one can understand and empathise with a great deal of Runciman's analysis.
The final chapters deal with the rise of AI and this part of the discussion is quite novel, as it incorporates the artificiality of the functioning of AIs within his narrative about the first singularity occurring during the rise of the modern state. But even here he operates with terms like ‘better answers’ that are ultimately very normative, even if he does not provide full answers or prescriptions about the better organisation of society. Rather, he considers what various mixes of Human Intelligence (HI) – Artificial Intelligence (AI) – and Artificial Agency (AA) mean.
While the book may be billed as a discourse on the future of AI, its primary objective is to uphold the enduring principles of an elite-centred variation of liberal ideology. It advocates for embracing the inherent detachment from centres of authority, reassuring individuals that such detachment serves their best interests. That is not to say that I did not enjoy reading it.
Read
September 28, 2024Wonderfully clear prose but requires concentration. Original analysis which likens our increasing reliance upon computer technology with people's increasing reliance upon other non-human institutions over the last 300 years (such as The State, limited companies and global corporations).
Want to read
December 5, 2023DMPL AUDIBLE EXAMINED 2023_12_05 Primary focus is AI - a good one to listen to. Focus on corporations and states as algorithmic entities is useful
February 25, 2024
Somewhat difficult to focus on and repetitive themes but Runciman makes his point .. I skimmed portions of the book when I failed to understand exactly what I was reading.
December 18, 2023
I envy anyone encountering Runciman for the first time. His podcast, “History of Ideas”, a series of lectures about historical ideas and essays, provoke the written word and well developed idea can be revolutionary.
One hardly blinks when confronted with the thought that we are living through the transformation of our century. The ascent of a non-human intelligence with the power to threaten, coerce, enlighten, and ultimately control. Runciman looks back at the artificial entities that were developed post enlightenment - empires and corporations. Human flourishing occurred in a way that was profound and singular, so much that the time period is called “The Great Transformation”. The refinement of the humanistic ideals from philosophers and jurisprudence have continued to be built and defended by those fortunate to live with rights and some autonomy.
We’ve had analogies of the mechanical embrace from a powerful sovereign in Hobbes “Leviathan”, or far reaching madness of science without discipline with Mary Shelly’s Frankenstein. Runciman goes on to explore human thinking - through duplicity in roles of leadership, groupthink, laws and challenges. The uneasy relationship of citizen and state is explored through generations of conflict.
The challenge of A.I. is different. The clarion call for disruption from our tech leaders may ultimately lead to a surpassing intelligence. A self-organizing AI could have the potential to replace jobs, creativity, government, economy and core relationships. Our identity as a species may be unrecognizable. As Runciman states, we have moved from lives of historically abysmal lives to lives of modern absurdity. The beast of human sapiens may further cosseted and removed by the creation of Yuval Noah Harari’s idea of Homo Dues.
All and all a really great read. Very dense and layered, and I hope to have another go at re-reading or at least learn more from Runciman’s lectures. One of the better AI reads i’ve come across that explores humanity’s fragile balancing of technology, human rights and a path toward a future of our choosing.
One hardly blinks when confronted with the thought that we are living through the transformation of our century. The ascent of a non-human intelligence with the power to threaten, coerce, enlighten, and ultimately control. Runciman looks back at the artificial entities that were developed post enlightenment - empires and corporations. Human flourishing occurred in a way that was profound and singular, so much that the time period is called “The Great Transformation”. The refinement of the humanistic ideals from philosophers and jurisprudence have continued to be built and defended by those fortunate to live with rights and some autonomy.
We’ve had analogies of the mechanical embrace from a powerful sovereign in Hobbes “Leviathan”, or far reaching madness of science without discipline with Mary Shelly’s Frankenstein. Runciman goes on to explore human thinking - through duplicity in roles of leadership, groupthink, laws and challenges. The uneasy relationship of citizen and state is explored through generations of conflict.
The challenge of A.I. is different. The clarion call for disruption from our tech leaders may ultimately lead to a surpassing intelligence. A self-organizing AI could have the potential to replace jobs, creativity, government, economy and core relationships. Our identity as a species may be unrecognizable. As Runciman states, we have moved from lives of historically abysmal lives to lives of modern absurdity. The beast of human sapiens may further cosseted and removed by the creation of Yuval Noah Harari’s idea of Homo Dues.
All and all a really great read. Very dense and layered, and I hope to have another go at re-reading or at least learn more from Runciman’s lectures. One of the better AI reads i’ve come across that explores humanity’s fragile balancing of technology, human rights and a path toward a future of our choosing.
April 7, 2025
I preferred the film. Instead of the Ed Helms character there's some guy named Thomas Hobbes and in place of the comedic tale of a group of friends trying to account for a forgotten night if excess involving a tiger, a baby and a stolen police car, we get a plot that encompasses basically everything that has ever happened. It was just too much to get my head around. I enjoy the author's podcast, but he's usually making fairly tight points in that. Here he is comparing various human created institutions and intelligence such as generative AI, the corporation and the state, and how they all interact with the aims of the wider human race, and I just didn't feel like there was enough of a clear thesis being put across it just seemed like he was trying to offer a grand theory of everything, coming from all sides at once on every level at once and I just couldn't follow it all, or even picture Zach Galifianakis in any of the lead roles.
May 20, 2024
Læst på dansk.
I modsætning til fx Rune Lykkeberg (https://e-avis.information.dk/titles/...) og David Trads i Jyllandsposten var jeg ikke begejstret. Ok, det er da muligvis interessant at blive opdateret på Hobbes og Leviathan som Runciman bruger mange sider på i begyndelsen af bogen, men pointen med at stater og selskaber er maskiner (?) og udtryk for kunstig intelligens (?) trængte ikke rigtig ind hos mig. Og at Runciman skulle efterlade læseren med håb, blev ikke tydeligt for mig.
Jeg synes bogen er lang, unødigt omstændelig og fyldt med idelige gentagelser og "...men på den anden side". For mig ville et godt langt essay have været at foretrække.
I modsætning til fx Rune Lykkeberg (https://e-avis.information.dk/titles/...) og David Trads i Jyllandsposten var jeg ikke begejstret. Ok, det er da muligvis interessant at blive opdateret på Hobbes og Leviathan som Runciman bruger mange sider på i begyndelsen af bogen, men pointen med at stater og selskaber er maskiner (?) og udtryk for kunstig intelligens (?) trængte ikke rigtig ind hos mig. Og at Runciman skulle efterlade læseren med håb, blev ikke tydeligt for mig.
Jeg synes bogen er lang, unødigt omstændelig og fyldt med idelige gentagelser og "...men på den anden side". For mig ville et godt langt essay have været at foretrække.
September 7, 2025
It was an interesting read and useful for orienting your head around the economic environment that has allowed businesses promoting AI products and services to grow.
Ultimately, I felt it was a little bloated for the amount of information provided. Far more concerned with the citiy and state portions of the book than AI speculation. This is perhaps due to a hesitation to indulge in educated speculation of events to come, instead preferring to recount historical precedence.
Ultimately, I felt it was a little bloated for the amount of information provided. Far more concerned with the citiy and state portions of the book than AI speculation. This is perhaps due to a hesitation to indulge in educated speculation of events to come, instead preferring to recount historical precedence.
April 25, 2024
Really great way of understanding the links between states and corporations (although perhaps only one way of doing that) and the artificiality of states and corporations. I take issue with the way he framed a few things like the Irish ‘famine’ but his overall argument is strong and really enlightening.
Recommend to everyone!
Recommend to everyone!
January 29, 2025
The best book I've ever read to explain why we are at this particularly challenging place in society, both politically and technologically and what choices lay before us. An outstanding book that breaks down the way that the construction of corporations and political governments has paved the way to inform how we think about handing decisions off to artificial intelligence.
November 8, 2023
I enjoyed this book! Very interesting take on state and corporation and people. I learned a lot from this book! An absolute delight to read.
Blog: http://literaturesandmovies.com
Blog: http://literaturesandmovies.com
December 14, 2023
I’m glad this book exists, we seem to be hurtling so quickly towards more and more advanced AI in our lives, and this book raises important questions about what that could all mean.
February 14, 2024
Some very interesting ideas, but I found it far too wordy.
Displaying 1 - 28 of 28 reviews