I got this book as it was recommended by the Financial Times for their books of 2021. AI 2041 is jointly written by Kai – Fu Lee, president of Google China from 2005 – 2009, and Chen Qiufan (alias Stanley Chan), also a former Google employee who ended up becoming a Sci – Fi writer. As they tell it in the book’s introduction, they are both excited about the prospect of AI and wish to write a book that would help explain where the concept of AI is in 2021, and how AI might help shape the world of 2041 – twenty years later. The book contains ten stories, each written by Chen Qiufan (in Chinese, translated into English), while Kai – Fu Lee provides a technical look at the technology highlighted in each story in a post-story analysis (written directly in English).
The first of the ten stories is “The Golden Elephant”, which highlights deep learning and big data, the fundamental breakthrough driving most of the AI innovations in the book. As such this story about a young girl in Mumbai finding love, only to have the AI attempt to steer her clear of it, is told first. The later stories build up on this foundation. For example, the next two stories (Gods Behind The Mask and Twin Sparrows) show the applications of deep learning, the former story showing how it can make deep fakes and the latter showing how it can act as a digital teacher for children by adapting to each child’s needs. The rest of the stories similarly build upon the foundations laid by the earlier stories to form a complete picture of 2041 as the two authors envision it.
In general, for the fiction parts, I feel there is a disconnect somewhere. As mentioned, Chen Qiufan wrote the stories in Chinese (Mandarin, I’m assuming) which was later translated into English. Either something is lost in translation, or there is an idiosyncrasy to the original language that is causing it – or both. Many stories have a lot of telling, which may be how Chinese fiction is written. I can’t say for sure, but it is jarring. Normally, when reading Western literature (especially modern literature) the common wisdom is to ‘show, don’t tell’ which results in the ‘camera’ point of view storytelling which I have gotten used to.To read stories in this way is much like reading a work written in a language you’ve only just learned, despite it being translated already. For this reason, the earlier stories don’t resonate as well as they probably should, and thus feel like weaker entries compared to the latter tales, which is a shame.
Still, once this little bump has been traversed, it’s easy to see why Chen Qiufan is President of the World Chinese Science Fiction Association – the stories are engaging, characters may be less rounded but still never one note, and most importantly for a book like this, the sci-fi elements (AI, in this case) take center stage without overshadowing the human elements. Chen Qiufan set out to write stories that showcase the world of 2041 after AI has taken root in our society, and I daresay he’s succeeded admirably. Kai Fu Lee’s analysis at the end of each story doesn’t have the same level of prose, but it sacrifices poesy for clarity, which is appreciated. He outlines the tech used in each story, their development and implications for their impact on society. I appreciate Mr Lee insisting on only using technology that either exists today or can be extrapolated reasonably from today’s level to where it might be in the future. This lends a ‘hardness’ to Mr Chen’s sci – fi, making it less ‘Star Trek’ and more ‘Interstellar’. Admirably, both co – authors never leave out the human element in their writings, Chen placing believable characters in his fiction and Kai Fu factoring in the human element (even, in one chapter, going as far as including philosophy – more on this later) in his analysis. Taken together, their works both entertain and inform, and leave the reader with many questions as to what the future may hold – the Holy Trinity of science fiction writing.
I’m always blown away by the creativity shown in science fiction, nevermind the hardness. Whether it features faster than light travel and laser swords that just…stop…in mid air, or if it features technology that both exists and could conceivably exist in the present day, the implications of the featured technologies have been stunning, and the AI presented in AI 2041 is no exception. Autonomous vehicles, X-Reality, even DeepFake masks – it’s all incredible to think about, and considering the impact the smartphone has had on society, it does make one wonder how these new technologies will affect us. The stories they manage to come up with are almost always interesting, too.
It’s hard to pick a favorite among the ten stories. For one, my opinion was biased because as I mentioned earlier the writing style wasn’t what I was used to, and it was only after making it past the stumbling block mentioned was I able to truly sink my teeth into the tales, and I was into story number five or six by that point. Second, the story and the analysis both from a cohesive whole, and I couldn’t just pick one story as a favorite without considering Kai Fu Lee’s analysis. But… if I had a gun to my head, I would pick “The Holy Driver” or “The Isle of Happiness” as my favorites.
The sixth story, The Holy Driver, tells of a Sri Lankan video gamer who has been recruited to act as an assistant driver for autonomous vehicles. According to Kai Fu’s analysis, he tells us that there are six steps classified by the Society of Automotive Engineers into achieving fully autonomous vehicles, from the first step (L0) which is no automation to the final step, (L5) in which the vehicle is fully autonomous. The technology shown in the story is between L4 and L5, with our protagonist called to drive the autonomous vehicles during emergencies in an augmented reality cockpit. This story is my favorite as I think this is where the sci fi truly takes off – the fiction is engaging, tells an amazing character driven story, and the analysis helps inform the tech shown in the story incredibly well. I believe this story is where the book is at its zenith, achieving the Holy Trinity I mentioned.
The ninth story, The Isle of Happiness, is a close second because it does something unexpected in the context of the book and what it claimed to set out to achieve initially – it tackles philosophy. A Russian billionaire is summoned to an island by a Qatari royal in order to test his algorithm in making people happy, and along the way various philosophies of happiness are considered from Maslow’s Pyramid to Eysenck’s Hedonistic Treadmill. It seems strange for a science fiction book to explore philosophy, but it works – throughout the book the authors have consistently tried to frame the effects of AI on the human. Mr Lee does bring up data protection and storage as technological aspects of his analysis, but even then most of what he tells us relates to the impact of data collection and the usage of these deep learning algorithms (I told you each story builds upon the first!) to create effects we, as users, could not have predicted. I rate this as one of my favorites because of that, but because it focuses less on the technological aspect I have to rate it lower than The Holy Driver. This is still a science fiction anthology and I was sold on learning about AI, so you’ve got to have some there, though the focus on philosophy and less ‘hard’ sciences is a welcome choice.
In fact, both the authors’ decision to highlight the human impact of AI on society is a brilliant one. Too many speculations of science fiction focus on the technology humanity can achieve without considering its effects on society, creating a sort of Tech Fetishism that worships the inexorable march of science. As a result, we can see in our society how technology has grown beyond humanity – without learning how to properly communicate with each other, we have given society the method to contact each and every single person on the planet, which has led to culture wars, the entrenchment of ideas we thought would disappear, and various ugliness of humanity on full display. Without learning how to properly process information, we have had information blasted to us from a fire hydrant, leading to fake news, ignorance triumphing over expert opinion and the spread of propaganda done unknowingly by the population. Science must always consider the potential negative impact on society of every new idea, or we risk plunging ourselves into a new dark age of our own making.
In short, AI 2041 did what I think its authors set out to. I am less ignorant of AI than before I read the book, and am incredibly excited to see the advances that they have put forward. I am also cognizant of the threats AI could pose to society, and hope that the world manages to implement failsafes to prevent them from coming true. I’d picked this book up in hopes of reading some amazing fiction. I was not disappointed, but I also got an extra treat – I learned something in the end.