What do you think?
Rate this book
An Artificial Revolution: On Power, Politics and AI
Privacy consultant and AI expert Ivana Bartoletti explores the urgent existential threat that artificial intelligence poses to international social justice When most people think about AI, they think about the future—from driverless cars and smart cities to HAL 9000 and the Terminator—and are unaware that AI is already creating a dystopian present. The third book in the Mood Indigo series argues that the rapid growth of the AI industry is threatening to undo decades of progress in human rights and global equality. AI has unparalleled transformative potential to reshape society, our economies and our working lives, but without legal scrutiny, international oversight and public debate, we are sleepwalking into a future written by algorithms which encode racist, sexist and classist biases into our daily lives. This book exposes the reality of the AI revolution, from the low-paid workers who toil to train algorithms to recognize cancerous polyps, to the rise of techno-racism and techno-chauvinism and the symbiotic relationship between AI and right wing populism.
156 pages, Paperback
Published October 1, 2020
8 people are currently reading
141 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 - 14 of 14 reviews
September 7, 2024
This book attempts a basic introduction to AI and its influence on power dynamics and politics, but ultimately flounders on multiple grounds.
In the relatively short book, Bartoletti explains the emergence of AI, tackling different aspects considered important to look at. She described some horrifying activities around AI and generating/filtering data that I wasn't aware of. It was also very interesting to read more detailed about the surveillance method employed in China.
However, I found her arguments became quite repetitive and sometimes far-fetched. I thought linking sexism and populism to AI were important, but only discussing the impact AI specifically has on them seemed to be reaching. It's not as if populism is on the rise purely because of the sway of AI in social media, for instance. A shift in broader political values of the lower and middle classes and feelings of being ignored or left behind seem to me to be at the root of populism, something that is then exacerbated by the darker side of AI techniques. Similarly, the notion that sexism is maintained primarily by technology seems equally short-sighted, reducing a number of factors-culture, religion and politics-to technology alone. These issues undermined the arguments Bartoletti was trying to make.
I think an important point missing in this book is the fact that the global workings and consequences of AI are for most people (at least for now) intangible and therefore less relevant. Bartoletti made a comparison to nuclear weapons, asking why we're not protesting AI the same way as we protested nuclear weapons. The dangers of nuclear weapons are rather more straightforward and far easier to imagine when Chernobyl, nuclear tests, and the image of the mushroom cloud is still on our minds. It is much more difficult to convince people of the importance of our privacy and human dignity. Most will only realise the importance of these after it is too late and a crisis occurs or their own data is breached. Bartoletti did mention the lack of awareness, but only briefly explained that people are 'too focused on the implications of the existence of highly advanced robots in the future, rather than around the more urgent question of what AI is doing now' (p.100). A good start to a thought that should have been elaborated on.
The main issue with this book is likely it's size-it's a relatively short book that wants to do too much with very few words. Another issue, but which we can't fault the author for, is that the book is at this point already out of date with developments on AI.
In the relatively short book, Bartoletti explains the emergence of AI, tackling different aspects considered important to look at. She described some horrifying activities around AI and generating/filtering data that I wasn't aware of. It was also very interesting to read more detailed about the surveillance method employed in China.
However, I found her arguments became quite repetitive and sometimes far-fetched. I thought linking sexism and populism to AI were important, but only discussing the impact AI specifically has on them seemed to be reaching. It's not as if populism is on the rise purely because of the sway of AI in social media, for instance. A shift in broader political values of the lower and middle classes and feelings of being ignored or left behind seem to me to be at the root of populism, something that is then exacerbated by the darker side of AI techniques. Similarly, the notion that sexism is maintained primarily by technology seems equally short-sighted, reducing a number of factors-culture, religion and politics-to technology alone. These issues undermined the arguments Bartoletti was trying to make.
I think an important point missing in this book is the fact that the global workings and consequences of AI are for most people (at least for now) intangible and therefore less relevant. Bartoletti made a comparison to nuclear weapons, asking why we're not protesting AI the same way as we protested nuclear weapons. The dangers of nuclear weapons are rather more straightforward and far easier to imagine when Chernobyl, nuclear tests, and the image of the mushroom cloud is still on our minds. It is much more difficult to convince people of the importance of our privacy and human dignity. Most will only realise the importance of these after it is too late and a crisis occurs or their own data is breached. Bartoletti did mention the lack of awareness, but only briefly explained that people are 'too focused on the implications of the existence of highly advanced robots in the future, rather than around the more urgent question of what AI is doing now' (p.100). A good start to a thought that should have been elaborated on.
The main issue with this book is likely it's size-it's a relatively short book that wants to do too much with very few words. Another issue, but which we can't fault the author for, is that the book is at this point already out of date with developments on AI.
April 7, 2025
An alarming, cogent and Cassandra-like warning from back in 2020 about the failure of both government and society to be prepared for the onslaught on our freedoms and privacy heralded by Artificial "Intelligence". Bartoletti makes it clear that machine learning depends very much on who is setting the texts and providing the data, and highlights the linkages between that alleged "learning" (data strip-mining) and the authoritarian regimes that are popping up like stage-four cancers across the world.
A good read, if hardly a fun one.
A good read, if hardly a fun one.
May 4, 2024
There were some good arguments here and there but the book at large seemed unsubstantiated or repetetive. Some points were taken for granted and the reader was not given any supporting evidence for some of the other claims. All in all, the book was not intellectually stimulating and did not present new ideas - perhaps needs an update given that AI tech moves fast and the examples presented here are no longer up to date.
August 14, 2020
A wonderful introduction into an extremely pressing issue in these modern times. The author touches upon the politics inherent within AI, as well as the potential politics in the application of AI technologies.
While the depth of discussion is a bit limited, and the author makes occasional statements that require a level of previous establishment, the breadth of topics mentioned makes up for it greatly.
While the depth of discussion is a bit limited, and the author makes occasional statements that require a level of previous establishment, the breadth of topics mentioned makes up for it greatly.
February 26, 2025
What is AI actually for in the first place?
Is it to punish, or to help?
This essay is a non-technical, high-level exploration of the inherently political nature of AI and the data that feed it. The author argues persuasively that AI-promoters / speculators have so far successfully avoided any serious public debate on this critical issue, one particularly important if we are to maintain any of our hard-fought incremental gains on the beaches of restorative justice and social equality.
The power underpinning AI is data. Data that are extracted and exploited aggressively. Data that reflect historical and existing unequal power structures and biases of racism, sexism, ageism, and elitism of all stripes. These biases are all coded into algos, which are already impacting most of the global population today, for the worse.
Bartoletti is not anti-AI. She makes a compelling argument here that AI is more than mere technology. Because AI and the data that underpin it IS power, society must democratically (actual democracy, not the corrupt simulacra of so-called representative d) decide how and where we want to deploy it. We need an ongoing public discussion, at least on the level of dialogue and debate on nuclear weapon proliferation that occurred in the 80s (which won many largely unheralded victories). The control of individual data must be wrenched back from the rapacious hands of those who have captured it through strategic manipulation, exploitation, and outright theft.
Writing in 2020, post Brexit and Trump 1.0 but before the latest series of least-common-denominator right-wing populist victories around the so-called developed world, Bartoletti noted:
In Ch 4, AI as Control, Bartoletti outlines plans in the US for HARPA, which would monitor neuro-behaviours--an even more dystopian vision of the future than PKD's short story, "Minority Report"--where AI will 'predict' violence and send what would of course be killer dog-like robots to 'intervene' before the next mass shooting. Much simpler and easier than challenging the gun pushers and dealers in the first place.;->
Is it to punish, or to help?
This essay is a non-technical, high-level exploration of the inherently political nature of AI and the data that feed it. The author argues persuasively that AI-promoters / speculators have so far successfully avoided any serious public debate on this critical issue, one particularly important if we are to maintain any of our hard-fought incremental gains on the beaches of restorative justice and social equality.
The power underpinning AI is data. Data that are extracted and exploited aggressively. Data that reflect historical and existing unequal power structures and biases of racism, sexism, ageism, and elitism of all stripes. These biases are all coded into algos, which are already impacting most of the global population today, for the worse.
Bartoletti is not anti-AI. She makes a compelling argument here that AI is more than mere technology. Because AI and the data that underpin it IS power, society must democratically (actual democracy, not the corrupt simulacra of so-called representative d) decide how and where we want to deploy it. We need an ongoing public discussion, at least on the level of dialogue and debate on nuclear weapon proliferation that occurred in the 80s (which won many largely unheralded victories). The control of individual data must be wrenched back from the rapacious hands of those who have captured it through strategic manipulation, exploitation, and outright theft.
I think it is now time to redefine privacy around freedom from persuasion, even freedom from being nudged towards a particular choice. In a digital age, we should be able to claim the right to our own journey through life without the intrusion of algorithms. It is about dignity. The persuasion architecture that underpins the world of social media is a breach of our essential rights as human beings, our integrity, our true selves.
Writing in 2020, post Brexit and Trump 1.0 but before the latest series of least-common-denominator right-wing populist victories around the so-called developed world, Bartoletti noted:
One problem with the architecture of persuasion is that it benefits populists around the world. The messages they propagate can be fuelled online and served to exactly the right people thanks to the obscurity and complexity of algorithmic-driven advertising.
...
It is not my intention to blame social media for the rise of populism, but it is my intention to show that is has been perhaps the most effective tool of populists. Because, alongside stagnating incomes for many and the steady increase in inequalities on a global scale, the politics of hate and the use of AI-driven behavioural advertising and predictive technologies cannot and must not be ignored. Without challenging the proliferation of hate online, of fake news and disinformation, women will see old attitudes resurfacing, and this is a threat we must confront.
In Ch 4, AI as Control, Bartoletti outlines plans in the US for HARPA, which would monitor neuro-behaviours--an even more dystopian vision of the future than PKD's short story, "Minority Report"--where AI will 'predict' violence and send what would of course be killer dog-like robots to 'intervene' before the next mass shooting. Much simpler and easier than challenging the gun pushers and dealers in the first place.;->
January 20, 2025
“Now is the time for a revolution […] in how the forces that will shape our future will […] be shaped. AI is political, and it is gendered.” In An Artificial Revolution: On Power, Politics and AI, an extended essay by Ivana Bartoletti, she makes the case for careful and thoughtful regulation being needed in the current global race to further the capabilities of AI — already-worrisome — which stands to affect society in ways that will abet historic oppressions from which marginalised groups are only starting to break free. Notions of Artificial Intelligence and Machine Learning being dependent on / developed by several biased and prejudiced factors, people, data, environments etc is not a stretch, nor is it unfamiliar to contemporary tech discourse. In fact, much of Bartoletti’s argument — and maybe its greatest strength — derives from her insistence on challenging the data economy and its “sanctity” (indeed, her note that “Some insist that data is the new oil” reminded me of Niall Bourke’s clever, oh-so-chilling dystopian novel Line). As well as the necessity of escaping “internalized” ideas of the value of data, Bartoletti points out how so much tech cones from “a lot of exploitative and degrading human labour.” I wasn’t always convinced by ideas of “data violence”, though this may be something I need to untangle a bit more, semantically, while further digesting Bartoletti’s thorough book; her voice, insisting on protecting society from itself, is a vital one.
August 10, 2023
Bartoletti raises some interesting points, especially the concern that we are sleep walking into an AI world without wrapping the right controls around it, or even challenging what any instance of AI is for. As Bartoletti says, we let big tech firms extract data for years before addressing privacy issues, and we are repeating those errors (or omissions) now.
This short book does a good job of introducing what AI is and some of the concerns we should hold tight to. Ironically, for a short book, it is a little repetitive, with points being driven home by stating them again and again.
This short book does a good job of introducing what AI is and some of the concerns we should hold tight to. Ironically, for a short book, it is a little repetitive, with points being driven home by stating them again and again.
May 4, 2021
Important read on the AI revolution where increasingly personal data is used for political gains. The author throws light on the issue of women and disadvantaged communities being more at a loss from the current AI development practices and explores the need for stronger policies regarding AI development.
May 31, 2021
Short but intense education on the key issues around AI. Key insights on bias (stop calling it bias - it is racism and sexism and discrimination), the effect of AI on women (not good, as it reflects and reinforces existing power structures) and the need for ethics based regulation, likely led by the EU.
January 29, 2023
Bartoletti offers an insight into the increasingly worrying world of AI, exploring how politics is shaped by the rise of AI, what we should use AI for, and how women are discriminated against in the technology sector. I’d highly recommend this as the widespread panic and attention to AI becomes more prevalent in 2023.
December 23, 2020
Short and informative
I most enjoyed the details and examples given such as surveillance techniques being used in China. Appreciated the book wasn’t bloated and didn’t feel the need to follow the trend of most non fiction books having 250-400 words.
I most enjoyed the details and examples given such as surveillance techniques being used in China. Appreciated the book wasn’t bloated and didn’t feel the need to follow the trend of most non fiction books having 250-400 words.
January 6, 2021
Fantastic read. Some very important questions and points raised. Essential reading I think not only for anyone in the AI field, but anyone in the software development lifecycle.
December 20, 2022
Crucial read
February 25, 2025
important stuff. already a bit outdated but what can you do
Displaying 1 - 14 of 14 reviews