royvdb's review against another edition
5.0
Finally a nuanced book about AI. When I saw the title I expected another biased book. China vs. Silicone Valley, AI as solution or AI as danger. Everything is possible or there will never be general AI... But in all those discussions, the author really talks about the pro's and cons of both sides. It's also a broad view on all elements of AI: The technological, economical, social, political and even a very introspective existential chapter. Loved it.
gajeam's review against another edition
2.0
This book makes a few interesting points about Chinese entrepreneurship culture, but it’s weighed down by industry-friendly, thought-leadery bullshit. Lee mentions in the acknowledgments that it took him six months to write, and you can tell in the end product. The book is dominated by citation-free takes and excerpts from the author’s resume. Worth checking out a review, but not the book itself.
mishlist's review against another edition
4.0
A really engaging examination of the progress of AI, its uses and its impact. Lee covers its early pioneers, progress through Chinese and American businesses, the 4 types of AI, internet, business, perception and autonomous. It's easy to read and understand because Lee includes a nunber of specific examples of real life application, for example on the internet, where AI makes recommendations to keep users engaged, as well as imagined applications in the future, such as a supermarket which combines a profile of shoppers and, their preferences with sales and promotion data to prompt purchases, drones swarms which fight bushfires and individual learning pathways for primary and high school students.
The later chapters tackle the impact of AI on jobs, and Lee draws on Moravec's Paradox - that while algorithms can develop to mimic the intellectual capabilities of adult, it is far more difficult to give robots the motor skills of a toddler. Lee places different occupations in the context of AI's capabilities - that the jobs we will no longer need are the asocial, repetitive and data driven ones but we will still very much need people to do the social work of talking to each other. Lee also makes the point that AI is drawn to monopolies - the better the product, the more users, the more users, the more data, the more data there is, the better the product, leading to more users and data. This builds a large barrier to entry. While I enjoyed most chapters and Lee's observations, a critique I have is that Lee continually raises the cultural differes and the contexts which differentiate the Chinese and American states, for example, how Chinese people have been willing to sacrifice some privacy for convenience. But all through the book, Lee shys away from making a clear determination about how the ethical and social risks can be addressed, instead repeating it will depend on government decisions and indvidual countries in their cultural context to adapt, and that we should be open-minded to different approaches to AI governance from across the world. While that could be a book in itself, it doesn't make much sense not to analyse any aspect of this when it forms such a significant difference in the way China and the America have developed their approach to AI and new technologies. Lee is also thoughtful about the purpose and impact of UBI - suggesting that many entrepreneurs see it as a quick all encompassing fix which means they don't need to consider comprehensive social policy.
Rather than UBI, Lee advocates for a social investment stipend, a payment to those who invest their time in a 'kind, compassionate and creative society', intended to be part of a new social contract, one which values socially beneficial activity the way we currently reward economically productive activity. Lee points to caring roles, particularly unpaid caring roles, which suggest there is a clear gap between what we value economically and what is important to us as a society. This and Lee's other reflections about the actual worth and value of humans in a society made AI Superpowers a worthwhile read.
The later chapters tackle the impact of AI on jobs, and Lee draws on Moravec's Paradox - that while algorithms can develop to mimic the intellectual capabilities of adult, it is far more difficult to give robots the motor skills of a toddler. Lee places different occupations in the context of AI's capabilities - that the jobs we will no longer need are the asocial, repetitive and data driven ones but we will still very much need people to do the social work of talking to each other. Lee also makes the point that AI is drawn to monopolies - the better the product, the more users, the more users, the more data, the more data there is, the better the product, leading to more users and data. This builds a large barrier to entry. While I enjoyed most chapters and Lee's observations, a critique I have is that Lee continually raises the cultural differes and the contexts which differentiate the Chinese and American states, for example, how Chinese people have been willing to sacrifice some privacy for convenience. But all through the book, Lee shys away from making a clear determination about how the ethical and social risks can be addressed, instead repeating it will depend on government decisions and indvidual countries in their cultural context to adapt, and that we should be open-minded to different approaches to AI governance from across the world. While that could be a book in itself, it doesn't make much sense not to analyse any aspect of this when it forms such a significant difference in the way China and the America have developed their approach to AI and new technologies. Lee is also thoughtful about the purpose and impact of UBI - suggesting that many entrepreneurs see it as a quick all encompassing fix which means they don't need to consider comprehensive social policy.
Rather than UBI, Lee advocates for a social investment stipend, a payment to those who invest their time in a 'kind, compassionate and creative society', intended to be part of a new social contract, one which values socially beneficial activity the way we currently reward economically productive activity. Lee points to caring roles, particularly unpaid caring roles, which suggest there is a clear gap between what we value economically and what is important to us as a society. This and Lee's other reflections about the actual worth and value of humans in a society made AI Superpowers a worthwhile read.
milkywaymerchant's review against another edition
5.0
Wrong version, not audio.
Review:
If you want to understand the depth of what we're facing in the future as it applies to AI, Machine Learning, The rise and fall of empires, and in some respect COVID-19. Then this book is for you.
Review:
If you want to understand the depth of what we're facing in the future as it applies to AI, Machine Learning, The rise and fall of empires, and in some respect COVID-19. Then this book is for you.
jamrock's review
5.0
This was such a surprisingly pleasant and informative read. I think I expected this to have a slight doomsday "impending AI cold war" feel to it but instead I was treated to a book which doesn't relentlessly pursue one argument about the race for AI supremacy but instead weaves a broad tapestry of our AI-augmented world, taking us through various related histories to where we are now before making some informed predictions about the future.
Dr Lee is from Taiwan but his career history spans many years in both the US and China, working as a senior executive in Microsoft, Apple and Google. Some takeaways from this book were:
* the brief history of deep learning, from when it was a poor relation of its knowledge-based sibling, through to its current meteoric, data fuelled rise.
* peering behind the curtain at China's alternate world of Internet and rise of the super apps like WeChat that were both fuelled by AI culture and also accelerants. This is coupled with an insiders view of Silicon Valley start-ups and the rise of VC funding in both countries. I guarantee you will never look at China's Internet the same way after reading this book.
* the clear explanation of the four waves of AI (where we are, I guess, cresting wave three) and a brief exploration of the utopian/dystopian supreme goal of "artificial general intelligence" which when taken to a logical extreme is actually the plot of the Arc of a Scythe series I recently finished reading.
Last but not least a truly inspiring last third of a book that takes a deeply humanistic look at how instead of AI "taking all the jobs" it could free people's time to focus on social projects that robots will never be able to emulate.
A non-fiction book with a feel good ending, who would have predicted it?
Dr Lee is from Taiwan but his career history spans many years in both the US and China, working as a senior executive in Microsoft, Apple and Google. Some takeaways from this book were:
* the brief history of deep learning, from when it was a poor relation of its knowledge-based sibling, through to its current meteoric, data fuelled rise.
* peering behind the curtain at China's alternate world of Internet and rise of the super apps like WeChat that were both fuelled by AI culture and also accelerants. This is coupled with an insiders view of Silicon Valley start-ups and the rise of VC funding in both countries. I guarantee you will never look at China's Internet the same way after reading this book.
* the clear explanation of the four waves of AI (where we are, I guess, cresting wave three) and a brief exploration of the utopian/dystopian supreme goal of "artificial general intelligence" which when taken to a logical extreme is actually the plot of the Arc of a Scythe series I recently finished reading.
Last but not least a truly inspiring last third of a book that takes a deeply humanistic look at how instead of AI "taking all the jobs" it could free people's time to focus on social projects that robots will never be able to emulate.
A non-fiction book with a feel good ending, who would have predicted it?
vinayvallala's review against another edition
4.0
This book gives a completely new perspective to the Chinese entrepreneurship. The author describes the AI development in China as sophisticated and may even beat silicon Valley in medium term. Then author describes possible disruptions on global scale due to AI and presents a way forward, which doesn't necessarily be UBI. The second half of the book is more like what world(dystopian way) probably looks like in 20-30 years and how to make sure that we won't reach that point.
Overall its a good read,which has lots of infotainment value.
Overall its a good read,which has lots of infotainment value.
rick2's review against another edition
4.0
Thought this would be a bit fluffy and light on content due to the cover and hyperbolic positioning. Glad to say this is a fantastic book from someone who is deeply immersed in the subject. Kai-Fu Lee worked for Microsoft’s Research Asia division in the 90s, built out much of Google China in the 2000’s and now does investment in the AI space in China.
As a Chinese investor in mostly Chinese companies, it follows that this is a mostly pro-China tech industry book. Basically every early chapter is oriented around “Why China will lead the world in this area.” For example in chapter 5 when talking about business-based intelligence and how US firms rely much more on external consulting which sets them up to have better data available, Lee ends his point by saying “but you never know, with technology anyone can leapfrog the current system” ok dude. I’ll buy a lot of it. But not all of it.
The book also neatly skips over delicate political subjects like the use of AI in northwestern China as a sort of panopticon technological prison. This book mostly focuses on how much potential the Chinese market has to be the global leader in the space, and often points to leading statistics on how much Chinese companies and universities are contributing. And while this book is not necessarily set up to critique the space that way it was a glaring omission.
However that being said, this book does a fantastic job of pointing to the advantages of China in the arms race that is AI research. Government support, national alignment and work ethic, and the growing education capabilities.
The book also takes a hard turn about halfway through into a more speculative area. What do we do about the changes coming due to this technology? The credibility built by the author pays off in a way that I think proved very effective. His personal story lends some credibility to his slightly “namaste” peace-love-and-happiness, take on how we should deal with the changes. I think his criticism of UBI inside is one of the better I’ve read. Positioning it as a sweeping “one-size-fits-all“ solution to a highly varied problem is flawed. And anyone with half a brain cell can agree that we need to compensate emotional labor and care positions in a more sustainable way.
I think many of the points are well-founded. Lee is obviously very enthusiastic at times and can get carried away with the subject he talks about, but that’s a given and easily understandable given his background. There’s some flaws that critical thinking solves easily, but overall this is one of the better books I’ve read about the topic.
I think anyone interested in AI and learning more about the next 5 to 20 years, or looking for insight into the Chinese technology market should read this book.
As a Chinese investor in mostly Chinese companies, it follows that this is a mostly pro-China tech industry book. Basically every early chapter is oriented around “Why China will lead the world in this area.” For example in chapter 5 when talking about business-based intelligence and how US firms rely much more on external consulting which sets them up to have better data available, Lee ends his point by saying “but you never know, with technology anyone can leapfrog the current system” ok dude. I’ll buy a lot of it. But not all of it.
The book also neatly skips over delicate political subjects like the use of AI in northwestern China as a sort of panopticon technological prison. This book mostly focuses on how much potential the Chinese market has to be the global leader in the space, and often points to leading statistics on how much Chinese companies and universities are contributing. And while this book is not necessarily set up to critique the space that way it was a glaring omission.
However that being said, this book does a fantastic job of pointing to the advantages of China in the arms race that is AI research. Government support, national alignment and work ethic, and the growing education capabilities.
The book also takes a hard turn about halfway through into a more speculative area. What do we do about the changes coming due to this technology? The credibility built by the author pays off in a way that I think proved very effective. His personal story lends some credibility to his slightly “namaste” peace-love-and-happiness, take on how we should deal with the changes. I think his criticism of UBI inside is one of the better I’ve read. Positioning it as a sweeping “one-size-fits-all“ solution to a highly varied problem is flawed. And anyone with half a brain cell can agree that we need to compensate emotional labor and care positions in a more sustainable way.
I think many of the points are well-founded. Lee is obviously very enthusiastic at times and can get carried away with the subject he talks about, but that’s a given and easily understandable given his background. There’s some flaws that critical thinking solves easily, but overall this is one of the better books I’ve read about the topic.
I think anyone interested in AI and learning more about the next 5 to 20 years, or looking for insight into the Chinese technology market should read this book.
themondello's review against another edition
5.0
Best book about AI with knowledge and personal perspective of author (which is very different and interesting). Highly recommend