What do you think?
Rate this book
The New Empire of AI: The Future of Global Inequality
As AI takes hold across the planet and wealthy nations seek to position themselves as global leaders of this new technology, the gap is widening between those who benefit from it and those who are subjugated by it. As Rachel Adams shows in this hard-hitting book, growing inequality is the single biggest threat to the transformative potential of AI. Not only is AI built on an unequal global system of power, it stands poised to entrench existing inequities, further consolidating a new age of empire.
AI’s impact on inequality will not be experienced in poorer countries it will be felt everywhere. The effects will be seen in intensified international migration as opportunities become increasingly concentrated in wealthier nations; in heightened political instability and populist politics; and in climate-related disasters caused by an industry blind to its environmental impact across supply chains.
We need to act now to address these issues. Only if the current inequitable trajectory of AI is halted, the incentives changed and the production and use of AI decentralized from wealthier nations will AI be able to deliver on its promise to build a better world for all.
AI’s impact on inequality will not be experienced in poorer countries it will be felt everywhere. The effects will be seen in intensified international migration as opportunities become increasingly concentrated in wealthier nations; in heightened political instability and populist politics; and in climate-related disasters caused by an industry blind to its environmental impact across supply chains.
We need to act now to address these issues. Only if the current inequitable trajectory of AI is halted, the incentives changed and the production and use of AI decentralized from wealthier nations will AI be able to deliver on its promise to build a better world for all.
224 pages, Hardcover
Published January 21, 2025
13 people are currently reading
369 people want to read
About the author
Rachel Adams
73 books10 followersRachel Adams is a writer and Professor of English and American Studies at Columbia University. She is the author of numerous academic articles and book reviews, as well as two books: Sideshow U.S.A.: Freaks and the American Cultural Imagination and Continental Divides: Remapping the Cultures of North America (both published by the University of Chicago Press). Her writing has also appeared in the Chronicle of Higher Education and the Times of London. Her book, Raising Henry: A Memoir of Motherhood, Disability, and Discovery, will be published by Yale University Press in September. She lives in New York City.
Source: About Rachel Adams.
Source: About Rachel Adams.
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 9 of 9 reviews
August 3, 2025
Sam Altman came to the University of Michigan in 2024 for a “fireside chat” with the dean of engineering. This event was coordinated because of the university’s rapid (and contentious) commitment to providing AI tools like chat GPT to university students free of charge. This step was taken while other universities were debating the ethics of AI in education and de-skilling that may result.
The event coincided with the release of GPTo, the first “reasoning” model from OpenAI. Sam was asked innocuous pre-screened questions about AI, and the event was largely a horse and pony show. Altman made the point that AI has the potential to revolutionize education for the better, allowing countries in the global developing majority to rapidly catch up with the industrialized world. This claim is made frequently, and at the time, I found it dubious at best and deceptive at worst. If AI has the power to revolutionize industry and education, won’t those with access to it further outpace the deeply impoverished if they cannot get their hands on the hardware required to interact with AI? (Smart phones, internet access, enormous supplies of reliable power). Now Sam might respond that smartphone ownership isn’t OpenAI’s problem, but then he can’t tout the potential benefit.
The optimistic view of an AI utopia is lauded as an altruistic tide that will raise all ships. This book calls that view into question, specifically focusing on the impact of AI in the “majority world” (third world, developing, global south, you get the idea). A post-colonial point of view is employed to argue that the majority world’s resources are being used to create AI, leaving them poorer, and they are not benefitting from the technology that results, as access is unequal and not curated to their hyperlocal, nonwestern viewpoints.
The argument can be divided into three major points.
1. Access to AI will not be equitable, and wealthier places that are developing it will only outpace their post-colonial holdings. Meanwhile, people in the majority world are suffering (and will only suffer more) from the process of creating AI.
2. LLMs and other AI tools are created by Westerners trained on the English language corpus, doubly biasing the generated content. This disconnect makes the tools less useful to the small number in the majority world that can access them and further degrades their local cultures that are already under significant pressure to conform to global trends.
3. Where AI tools are accessible, their use has significant negative effects, like misinformation and the enabling of authoritarianism through mass surveillance and biometric tracking.
Largely, this is a continuation of the colonial trend. Imperialist nations looted their colonies of their resources to industrialize, leaving them forever in the dust, unable to industrialize themselves as their former oppressors outpace them further and further. At present, AI tools are developed on hardware using mined materials from these same countries, tested on their populations, and used by their dictatorial leaders to maintain power. As a result, we have access to the most powerful technology ever created, while they are left behind. The same story is playing out, but now with exponential outpacing.
Here are some of the topics that are covered in support of each argument.
1. -Mining of lithium, cobalt and other materials from places like Chile and the Congo.
-Content moderators for the datasets used to train AI paid a pittance to watch and flag videos of violent mutilation, torture, and sexual violence and abuse for up to 12 hours a day. The moderators are often refugees in camps or poor and desperate people.
-Data centers degrade the quality of power grids due to the massive loads that draw power. If data centers are eventually built in third world countries, which is highly likely, their already unstable grids will degrade, harming people with already shoddy access to electricity.
2. AI tools are not useful to people who cannot use them. If the training data has no context of local cultures and local issues, it will likely provide misleading or unhelpful information. This is particularly true where literacy and English speaking rates are low. Even if translation to hyperlocal dialects is possible, meaning will be lost and local traditions harmed. Additionally, information heavily biased to western viewpoints will spread unchallenged, killing discourse on ethical and philosophical issues necessary for moral progress. Other AI tools, such as those used to determine access to loans, can subvert local customs and enforce racial biases.
3. The surveillance potential of AI is bone chilling. Already, biometric technologies are used by dictators to control and track their populations. These tools are being developed and tested in the majority world, and may eventually find their way back to the US once perfected. AI and social media perpetuates and exacerbate local conflicts, which destabilizes whole nations.
The positive potential of AI is undeniable, but the claim that it will help the poor goes largely unchallenged, though it is not untested. Historical evidence and early reports show that AI is not, in fact, making the world a better place. There are a million calls to action, but one that I would suggest is for CEOs of tech firms like Sam Altman to be publicly challenged when they make claims like those at the UofM fireside chat. How can AI close an economic rift when students at one of the wealthiest universities in the world are using an all powerful tool designed specifically for them, and people largely concerned with farming to feed their families don’t have access to their own version of these tools?
The event coincided with the release of GPTo, the first “reasoning” model from OpenAI. Sam was asked innocuous pre-screened questions about AI, and the event was largely a horse and pony show. Altman made the point that AI has the potential to revolutionize education for the better, allowing countries in the global developing majority to rapidly catch up with the industrialized world. This claim is made frequently, and at the time, I found it dubious at best and deceptive at worst. If AI has the power to revolutionize industry and education, won’t those with access to it further outpace the deeply impoverished if they cannot get their hands on the hardware required to interact with AI? (Smart phones, internet access, enormous supplies of reliable power). Now Sam might respond that smartphone ownership isn’t OpenAI’s problem, but then he can’t tout the potential benefit.
The optimistic view of an AI utopia is lauded as an altruistic tide that will raise all ships. This book calls that view into question, specifically focusing on the impact of AI in the “majority world” (third world, developing, global south, you get the idea). A post-colonial point of view is employed to argue that the majority world’s resources are being used to create AI, leaving them poorer, and they are not benefitting from the technology that results, as access is unequal and not curated to their hyperlocal, nonwestern viewpoints.
The argument can be divided into three major points.
1. Access to AI will not be equitable, and wealthier places that are developing it will only outpace their post-colonial holdings. Meanwhile, people in the majority world are suffering (and will only suffer more) from the process of creating AI.
2. LLMs and other AI tools are created by Westerners trained on the English language corpus, doubly biasing the generated content. This disconnect makes the tools less useful to the small number in the majority world that can access them and further degrades their local cultures that are already under significant pressure to conform to global trends.
3. Where AI tools are accessible, their use has significant negative effects, like misinformation and the enabling of authoritarianism through mass surveillance and biometric tracking.
Largely, this is a continuation of the colonial trend. Imperialist nations looted their colonies of their resources to industrialize, leaving them forever in the dust, unable to industrialize themselves as their former oppressors outpace them further and further. At present, AI tools are developed on hardware using mined materials from these same countries, tested on their populations, and used by their dictatorial leaders to maintain power. As a result, we have access to the most powerful technology ever created, while they are left behind. The same story is playing out, but now with exponential outpacing.
Here are some of the topics that are covered in support of each argument.
1. -Mining of lithium, cobalt and other materials from places like Chile and the Congo.
-Content moderators for the datasets used to train AI paid a pittance to watch and flag videos of violent mutilation, torture, and sexual violence and abuse for up to 12 hours a day. The moderators are often refugees in camps or poor and desperate people.
-Data centers degrade the quality of power grids due to the massive loads that draw power. If data centers are eventually built in third world countries, which is highly likely, their already unstable grids will degrade, harming people with already shoddy access to electricity.
2. AI tools are not useful to people who cannot use them. If the training data has no context of local cultures and local issues, it will likely provide misleading or unhelpful information. This is particularly true where literacy and English speaking rates are low. Even if translation to hyperlocal dialects is possible, meaning will be lost and local traditions harmed. Additionally, information heavily biased to western viewpoints will spread unchallenged, killing discourse on ethical and philosophical issues necessary for moral progress. Other AI tools, such as those used to determine access to loans, can subvert local customs and enforce racial biases.
3. The surveillance potential of AI is bone chilling. Already, biometric technologies are used by dictators to control and track their populations. These tools are being developed and tested in the majority world, and may eventually find their way back to the US once perfected. AI and social media perpetuates and exacerbate local conflicts, which destabilizes whole nations.
The positive potential of AI is undeniable, but the claim that it will help the poor goes largely unchallenged, though it is not untested. Historical evidence and early reports show that AI is not, in fact, making the world a better place. There are a million calls to action, but one that I would suggest is for CEOs of tech firms like Sam Altman to be publicly challenged when they make claims like those at the UofM fireside chat. How can AI close an economic rift when students at one of the wealthiest universities in the world are using an all powerful tool designed specifically for them, and people largely concerned with farming to feed their families don’t have access to their own version of these tools?
September 6, 2025
The New Empire of AI aims to highlight the risks of artificial intelligence exacerbating societal inequalities, drawing parallels to colonial structures and unchecked capitalism. While the premise, that AI could widen gaps and cause harm, has potential, the book delivers a muddled, one-sided argument that leans on ideological signaling rather than rigorous evidence or argument.
Fundamentally, the book struggles with a counter factual. Clearly, the author is well read on the areas of AI that suit their thesis, but readily available data on AI’s capacity to bridge divides between the haves and have-nots, was readily present in my mind as I read this.
Adams misses this obvious opportunity to strengthen her analysis by engaging with a clear counter and explaining why I should care more about her position or even why her position remains valid. The result is an unsatisfying read that left me frustrated and unconvinced. The core reflection through the book ended up because of this being “sure, but this seems to go the other way as well”.
Adams argues that AI, wielded by poorly incentivised actors, will perpetuate wealth disparities and harm marginalised groups. For example, her discussion of AI’s role in expanding the gig economy, eroding employment rights, and fueling the university essay scam industry feels selective and underdeveloped, lacking a counterfactual to show how AI’s impact differs from existing trends.
A glaring omission is the failure to address how AI can narrow inequalities, a perspective that would have been easy to incorporate given abundant evidence which she wouldn’t even have had to agree with to engage with. For instance, AI-powered tools have demonstrably expanded access to education, a critical driver of economic mobility. A 2023 UNESCO report notes that mobile learning platforms, enhanced by AI, have reached over 50 million learners in low-income countries, providing free or low-cost access to quality education resources. Platforms like Khan Academy and Coursera use AI to personalize learning, enabling students in remote or underserved areas to access materials once limited to wealthier regions. This directly counters Adams’ narrative of AI as a tool of exclusion, yet she ignores such examples.
Healthcare is another area where AI bridges gaps. AI-driven telemedicine platforms, such as those deployed in rural India and Sub-Saharan Africa, have improved access to diagnostics for millions. A 2022 World Health Organization study highlighted how AI tools for medical imaging analysis, like those detecting tuberculosis, have scaled healthcare delivery in regions with fewer than 1 doctor per 1,000 people. These tools are low-cost and portable, directly benefiting the “have-nots” by bypassing traditional barriers like distance or scarcity of skilled professionals.
Economic inclusion also sees gains from AI’s scalability. Mobile banking platforms like M-Pesa, enhanced by AI-driven fraud detection and credit scoring, have empowered over 50 million users in Africa to access financial services, per a 2024 GSMA report. This has enabled small-scale entrepreneurs in low-income communities to secure loans and participate in markets previously dominated by the wealthy.
Such examples illustrate AI’s potential to level the playing field, yet Adams sidesteps them, focusing instead on pre-existing capitalist and/or colonial flaws without weighing AI’s transformative benefits.
Even at the most basic level - that this sees power concentrated in the hands of a few is contestable. The book recognises that AI existed in less scalable forms prior, held by major companies and investors (see: social media). Now, anyone with an internet connection (increasing) can access tools used by the same people.
Later in the book, Adams briefly acknowledges the limits of AI’s scalability but again fails to construct a meaningful counterfactual. She critiques AI’s potential harms without noting how its low-cost deployment could deliver services to the “majority world.” For instance, AI-driven language translation tools, like Google Translate’s advancements in low-resource languages, have made information accessible to millions of non-English speakers, fostering inclusion in global knowledge economies.
The New Empire of AI leans on familiar left-liberal tropes: decrying capitalism, colonialism, and inequality, without offering fresh insights or practical solutions. The book’s refusal to engage with AI’s democratising potential feels like a deliberate oversight. These examples were not hard to find; a cursory review of global development reports or tech impact studies would have sufficed. I did it off reading about current events and googling while reading this. Most of the data falls within the books time period as well.
By ignoring them, Adams delivers a shallow critique that fails to advance a nuanced understanding of AI’s risks and opportunities.
Fundamentally, the book struggles with a counter factual. Clearly, the author is well read on the areas of AI that suit their thesis, but readily available data on AI’s capacity to bridge divides between the haves and have-nots, was readily present in my mind as I read this.
Adams misses this obvious opportunity to strengthen her analysis by engaging with a clear counter and explaining why I should care more about her position or even why her position remains valid. The result is an unsatisfying read that left me frustrated and unconvinced. The core reflection through the book ended up because of this being “sure, but this seems to go the other way as well”.
Adams argues that AI, wielded by poorly incentivised actors, will perpetuate wealth disparities and harm marginalised groups. For example, her discussion of AI’s role in expanding the gig economy, eroding employment rights, and fueling the university essay scam industry feels selective and underdeveloped, lacking a counterfactual to show how AI’s impact differs from existing trends.
A glaring omission is the failure to address how AI can narrow inequalities, a perspective that would have been easy to incorporate given abundant evidence which she wouldn’t even have had to agree with to engage with. For instance, AI-powered tools have demonstrably expanded access to education, a critical driver of economic mobility. A 2023 UNESCO report notes that mobile learning platforms, enhanced by AI, have reached over 50 million learners in low-income countries, providing free or low-cost access to quality education resources. Platforms like Khan Academy and Coursera use AI to personalize learning, enabling students in remote or underserved areas to access materials once limited to wealthier regions. This directly counters Adams’ narrative of AI as a tool of exclusion, yet she ignores such examples.
Healthcare is another area where AI bridges gaps. AI-driven telemedicine platforms, such as those deployed in rural India and Sub-Saharan Africa, have improved access to diagnostics for millions. A 2022 World Health Organization study highlighted how AI tools for medical imaging analysis, like those detecting tuberculosis, have scaled healthcare delivery in regions with fewer than 1 doctor per 1,000 people. These tools are low-cost and portable, directly benefiting the “have-nots” by bypassing traditional barriers like distance or scarcity of skilled professionals.
Economic inclusion also sees gains from AI’s scalability. Mobile banking platforms like M-Pesa, enhanced by AI-driven fraud detection and credit scoring, have empowered over 50 million users in Africa to access financial services, per a 2024 GSMA report. This has enabled small-scale entrepreneurs in low-income communities to secure loans and participate in markets previously dominated by the wealthy.
Such examples illustrate AI’s potential to level the playing field, yet Adams sidesteps them, focusing instead on pre-existing capitalist and/or colonial flaws without weighing AI’s transformative benefits.
Even at the most basic level - that this sees power concentrated in the hands of a few is contestable. The book recognises that AI existed in less scalable forms prior, held by major companies and investors (see: social media). Now, anyone with an internet connection (increasing) can access tools used by the same people.
Later in the book, Adams briefly acknowledges the limits of AI’s scalability but again fails to construct a meaningful counterfactual. She critiques AI’s potential harms without noting how its low-cost deployment could deliver services to the “majority world.” For instance, AI-driven language translation tools, like Google Translate’s advancements in low-resource languages, have made information accessible to millions of non-English speakers, fostering inclusion in global knowledge economies.
The New Empire of AI leans on familiar left-liberal tropes: decrying capitalism, colonialism, and inequality, without offering fresh insights or practical solutions. The book’s refusal to engage with AI’s democratising potential feels like a deliberate oversight. These examples were not hard to find; a cursory review of global development reports or tech impact studies would have sufficed. I did it off reading about current events and googling while reading this. Most of the data falls within the books time period as well.
By ignoring them, Adams delivers a shallow critique that fails to advance a nuanced understanding of AI’s risks and opportunities.
December 6, 2025
Dr. Rachel Adams presents a critical assessment of the uneven global landscape in which artificial intelligence is being developed and deployed. She contends that AI, embedded within pre-existing geopolitical and economic asymmetries, has the potential to exacerbate global inequality and entrench new forms of digital imperialism. The analysis highlights how these dynamics may influence migration patterns, political stability, and environmental outcomes, particularly in regions such as Sub-Saharan Africa. Adams ultimately argues for the urgent need to decentralise AI production and reform incentive structures to ensure that its benefits are equitably realised.
Key focus areas for AI Engineers in Africa
1. Build AI that Solves Local Problems, Not Imported Ones
Africa’s challenges—and opportunities—are distinct. AI engineers should focus on solutions tailored to local contexts such as agriculture, financial inclusion, healthcare access, language technologies, and climate resilience. Local relevance is a strategic advantage.
2. Prioritise Data Sovereignty and Ethical Data Practices
Much of Africa’s value in the AI ecosystem will come from data. Protecting this data from exploitation is crucial. Engineers must champion responsible data governance, local data ownership, and transparent collection practices to avoid reinforcing external dependencies.
3. Invest in Low-Resource and Infrastructure-Aware AI
Many mainstream AI models assume abundant compute and connectivity—conditions not always available locally. Engineers should optimise for low-bandwidth, low-power environments and push innovation in efficient algorithms, edge computing, and resource-light architectures.
4. Build Skills and Communities That Break Dependency Cycles
Collaborative learning, open-source contribution, and regional research networks will be essential for building independent AI ecosystems. Strengthening local expertise reduces reliance on foreign platforms and fosters indigenous innovation.
5. Advocate for Inclusive and Fair AI Policy
Technical decisions are inseparable from governance. Engineers must engage with policymakers to ensure that national AI strategies promote equity, transparency, and protection from exploitation by global corporations and foreign interests.
6. Develop AI With Cultural and Linguistic Awareness
Africa’s linguistic and cultural diversity is often ignored by mainstream AI systems. Engineers should prioritise multilingual, culturally grounded AI models to ensure accessibility and representation for all communities.
7. Embrace Open-Source as a Pathway to Autonomy
Open-source tools and community-driven research can help African engineers circumvent high licensing costs and restrictive technologies. They offer a route to innovation without dependence on foreign tech monopolies.
8. Ensure AI Strengthens, Not Undermines, Social Equity
Engineers should be critically aware of how AI systems may reproduce historical inequalities. Designing for fairness, transparency, and accountability is essential to prevent digital systems from becoming tools of exclusion.
9. Prepare for the Geopolitical Implications of AI**
As global competition intensifies, Africa risks becoming a site of extraction—of data, talent, and influence. Engineers must recognize these dynamics and build technologies that reinforce local autonomy rather than foreign control.
---
## **10. Build With Long-Term Resilience, Not Short-Term Adoption**
Sustainable AI solutions must withstand political, economic, and infrastructural instability. Robustness, adaptability, and long-term community stewardship matter more than rapid deployment.
---
If you'd like, I can also prepare:
🌍 A **policy brief** for African governments
📌 A **1-page summary** for engineers
🎤 A **talk outline** for conferences or meetups
📚 A **curriculum map** for training AI engineers in Africa
Key focus areas for AI Engineers in Africa
1. Build AI that Solves Local Problems, Not Imported Ones
Africa’s challenges—and opportunities—are distinct. AI engineers should focus on solutions tailored to local contexts such as agriculture, financial inclusion, healthcare access, language technologies, and climate resilience. Local relevance is a strategic advantage.
2. Prioritise Data Sovereignty and Ethical Data Practices
Much of Africa’s value in the AI ecosystem will come from data. Protecting this data from exploitation is crucial. Engineers must champion responsible data governance, local data ownership, and transparent collection practices to avoid reinforcing external dependencies.
3. Invest in Low-Resource and Infrastructure-Aware AI
Many mainstream AI models assume abundant compute and connectivity—conditions not always available locally. Engineers should optimise for low-bandwidth, low-power environments and push innovation in efficient algorithms, edge computing, and resource-light architectures.
4. Build Skills and Communities That Break Dependency Cycles
Collaborative learning, open-source contribution, and regional research networks will be essential for building independent AI ecosystems. Strengthening local expertise reduces reliance on foreign platforms and fosters indigenous innovation.
5. Advocate for Inclusive and Fair AI Policy
Technical decisions are inseparable from governance. Engineers must engage with policymakers to ensure that national AI strategies promote equity, transparency, and protection from exploitation by global corporations and foreign interests.
6. Develop AI With Cultural and Linguistic Awareness
Africa’s linguistic and cultural diversity is often ignored by mainstream AI systems. Engineers should prioritise multilingual, culturally grounded AI models to ensure accessibility and representation for all communities.
7. Embrace Open-Source as a Pathway to Autonomy
Open-source tools and community-driven research can help African engineers circumvent high licensing costs and restrictive technologies. They offer a route to innovation without dependence on foreign tech monopolies.
8. Ensure AI Strengthens, Not Undermines, Social Equity
Engineers should be critically aware of how AI systems may reproduce historical inequalities. Designing for fairness, transparency, and accountability is essential to prevent digital systems from becoming tools of exclusion.
9. Prepare for the Geopolitical Implications of AI**
As global competition intensifies, Africa risks becoming a site of extraction—of data, talent, and influence. Engineers must recognize these dynamics and build technologies that reinforce local autonomy rather than foreign control.
---
## **10. Build With Long-Term Resilience, Not Short-Term Adoption**
Sustainable AI solutions must withstand political, economic, and infrastructural instability. Robustness, adaptability, and long-term community stewardship matter more than rapid deployment.
---
If you'd like, I can also prepare:
🌍 A **policy brief** for African governments
📌 A **1-page summary** for engineers
🎤 A **talk outline** for conferences or meetups
📚 A **curriculum map** for training AI engineers in Africa
August 26, 2025
“Artificial intelligence is the future not only of Russia but of all of mankind. There are huge opportunities, but also threats that are difficult to foresee today. Whoever becomes the leader in this sphere will become the ruler of the world.” Vladimir Putin, September 2017.
I was gifted a Kindle copy of the book published on January 21, 2025, in July 2025. I started reading it on the evening of Sunday, August 24, 2025, and finished the next day, August 25, 2025.
It took me about 12 hours of continuous reading. I really enjoyed the book, as it addressed my concerns about the handling of AI in Sub-Saharan Africa.
As AI continues to expand globally, wealthy nations are striving to establish themselves as leaders in this emerging technology. However, this is widening the gap between those who benefit from AI and those who are oppressed by it.
In her insightful book, Rachel Adams argues that growing inequality poses the greatest threat to AI's transformative potential. Not only is AI built on an unequal global power structure, but it also threatens to reinforce existing disparities, ushering in a new era of empire.
The impact of AI on inequality will be felt globally, not just in poorer countries. We will see intensified international migration as opportunities increasingly concentrate in wealthier nations. This situation will lead to heightened political instability and the rise of populist politics. Additionally, climate-related disasters will occur as industries remain unaware of their environmental impact across supply chains.
We need to take immediate action to address these issues. If we do not halt the current inequitable trajectory of AI, change the incentives, and decentralize the production and use of AI away from wealthier nations, we will not be able to realize its potential to create a better world for everyone.
I was gifted a Kindle copy of the book published on January 21, 2025, in July 2025. I started reading it on the evening of Sunday, August 24, 2025, and finished the next day, August 25, 2025.
It took me about 12 hours of continuous reading. I really enjoyed the book, as it addressed my concerns about the handling of AI in Sub-Saharan Africa.
As AI continues to expand globally, wealthy nations are striving to establish themselves as leaders in this emerging technology. However, this is widening the gap between those who benefit from AI and those who are oppressed by it.
In her insightful book, Rachel Adams argues that growing inequality poses the greatest threat to AI's transformative potential. Not only is AI built on an unequal global power structure, but it also threatens to reinforce existing disparities, ushering in a new era of empire.
The impact of AI on inequality will be felt globally, not just in poorer countries. We will see intensified international migration as opportunities increasingly concentrate in wealthier nations. This situation will lead to heightened political instability and the rise of populist politics. Additionally, climate-related disasters will occur as industries remain unaware of their environmental impact across supply chains.
We need to take immediate action to address these issues. If we do not halt the current inequitable trajectory of AI, change the incentives, and decentralize the production and use of AI away from wealthier nations, we will not be able to realize its potential to create a better world for everyone.
November 27, 2025
As other reviewers have already said better than myself, this is a “shallow critique” of the topic. Very buzzword heavy.
Also, as much as I enjoy being endlessly lectured at by the white, middle-class, expensively-educated lady who wrote this, she probably typed it up on her top-end Apple Mac laptop whilst slurping down endless Pumpkin Spice lattes!
Not a peep about *her* privilege though.
As Bill Burr once said “B*tch! You were in the hot tub next to us!”
Her proposed solutions are fairly naive and wouldn’t work. Technological development would presumably be slowed to an effective halt until every pile of sticks in Africa catches up.
The truth is that there are no easy answers to this. I don’t have any, which is why you won’t catch me writing a below-average book pushing my ideas at you.
She also does that thing where she says "Women who went through all the glass ceiling challenges were better than the men when they got there."
She fails to see that this is *because* they went through challenges and a glass ceiling. It filtered out all the average women!
If those weren't there then the women who reached to top would be the same useless, corrupt equivalent of their men counterparts who sabotaged the careers of the more skilled women because they don’t want the competition!
Power is what corrupts people. Not ‘being a man’.
Anyway, that’s something of a rant, and it’s hugely off-topic.
Tldr: I’m not literally Hitler. I found her book exasperating. It could have been so much better. I've also exaggerated my views in an attempt at comedic purposes.
Also, as much as I enjoy being endlessly lectured at by the white, middle-class, expensively-educated lady who wrote this, she probably typed it up on her top-end Apple Mac laptop whilst slurping down endless Pumpkin Spice lattes!
Not a peep about *her* privilege though.
As Bill Burr once said “B*tch! You were in the hot tub next to us!”
Her proposed solutions are fairly naive and wouldn’t work. Technological development would presumably be slowed to an effective halt until every pile of sticks in Africa catches up.
The truth is that there are no easy answers to this. I don’t have any, which is why you won’t catch me writing a below-average book pushing my ideas at you.
She also does that thing where she says "Women who went through all the glass ceiling challenges were better than the men when they got there."
She fails to see that this is *because* they went through challenges and a glass ceiling. It filtered out all the average women!
If those weren't there then the women who reached to top would be the same useless, corrupt equivalent of their men counterparts who sabotaged the careers of the more skilled women because they don’t want the competition!
Power is what corrupts people. Not ‘being a man’.
Anyway, that’s something of a rant, and it’s hugely off-topic.
Tldr: I’m not literally Hitler. I found her book exasperating. It could have been so much better. I've also exaggerated my views in an attempt at comedic purposes.
August 31, 2025
This book mostly disparages AI and after reading Empire of AI by Karen Hao some of the stories of injustice in Kenya and other parts of the world are repeated. This book has more on an African perspective,but I didnt feel like it was as immersive as Hao's book. The overall premise of the book is AI is a way extend Colonialism virtually and because that's being done mostly in English the benefits are said to accrue to the global north at the expense of the global south and for that matter,greater inequality and polarization are also the likely outcome of this technology as it's tentacles seek to augment everyone. I've listened to or read a great deal about this technology recently and feel that the amount of electricity needed and the clean water it demands may make it infeasible unless fusion energy can be harnassed. I found both Nexus and Empire of AI far more engaging and thus gave them higher ratings.
July 14, 2025
This was a very thoughtful work and brings up many issues we need to worry about at Ai changes the economic landscape and is concentrated in a very few hands. Aside from the drain on electricity and water these powerful new tools represent, which has also been discussed in Crawford's book, I think the discussion of how Africa is almost completely ignored in the training datasets for Ai is a unique contribution of this book to the conversation.
July 27, 2025
An amazing book that explains how AI is impacting inequality in the world, it gives examples and it highlights different part of the industry that are not showcased normally. I understood the materiality of AI and how the people are involved, it’s truly a must to read!
December 9, 2025
Excellent wake-up call. Well written. Privileges the perspective of vulnerable populations. 5/5.
Displaying 1 - 9 of 9 reviews