Being Human in the Age of Artificial Intelligence

Artificial Intelligence is analyzed here with the eye of the science.
Nobody knows exactly how and when this major iteration of life will happen, but everybody should be aware that it is going to happen.
Max Tegmark's work is a great contribution in this direction.

A brilliant book! I first read it over two months ago.. and thought it was engaging.. As I give it another shot for some research for a book on the same topic... the subtlety of the arguments really came through... I think this is an amazing writer writing an amazing book! Max Tegmark, none can thank you enough!

Its a very provoking discussion, not focused on placing one technique above other but instead drive the curious (or student) towards important (not less urgent) topics around the artificial intelligence discussion that otherwise will not be thought in advance.

Truly remarkable work about AI challenges ... this book is surely a reference for anyone involved in building AI systems as well as policy makers and any entity involved with the consequences of super intelligence.

This bad review is not because of the content but because of the printing quality. I can barely flick through the pages. Look at the photo attached. What the hell? Is this thing even authentic? Very disappointing.

Edit: I was told that the finishing is done on purpose (deckled edge). Apparently is something that some people would pay a higher price for. However, I am not happy with this printing finishing and I would be better of flipping the pages smoothly. There should be a mention of this somewhere in the description.

This book was very interesting. The author does a good job of getting the reader to the point at which she/he can make an informed guess on whether they think AI will become unwieldy, dangerous, or downright destructive. Fun read.

I thoroughly enjoyed the detailed look at all of the current theories and worries about future AI and how it could affect (or eliminate) humans.

I think the likelihood of a community of AI (like the Matrix) is more likely than seems to be discussed in the book.

I experienced a lot of ambivalence in reading this work. There is a lot of "what if" and (wild) conjecture - both toward the good and the ill in AI. I've worked in the AI domain for over 10 years, and one of the biggest problems is with the anthropomorphizing of "intelligence". Being a domain in its earliest development, may of the assumptions may prove to be wrong. Furthermore, many of the potential solutions to hypothetical problems were never explored. But at least people are starting to talk about a future where humans and AI (along a spectrum) co-exist, if not co-evolve.

Straight to the point, I would not recommend this. To me the title is very misleading and is only touched upon briefly. Life 3.0 as referenced (fleetingly) in the book is the upload of human consciousness into a computer lending the ability to theoretically live forever. To me, this is a very interesting topic that I would love to know more about even though I can't see it as a viable option. Instead this book is nothing but a lot of (useless) information about how intellectuals socialize. The author is obviously an Elon Musk fanboy and tosses his name in wherever vaguely applicable (or not). Don't waster your time; it's a painful read and you won't have learned anything about AI in the end.

Tegmark brings a refreshing perspective to what likely is one of, if not the, most important conversation of our time. After setting the stage and clearing the field of common myths and misconceptions regarding AI, Tegmark methodically moves through the emergence of intelligence in our cosmos some 4 billion years ago to the implications of what he calls "Life 3.0," entities which can both redesign their hardware and software.

Couching the emergence and development of both consciousness and intelligence within the more cosmological world view of a physicist offers a truly inspiring narrative. Moving through hundreds, thousands, and billions of years with Tegmark solidifies the development of intelligence as a cosmological phenomenon which you come to realize that you yourself are a part of. Learning how intelligence is an emergent expression of more basic physical laws feels, at least to me, thoroughly grounding and deeply reestablishing of a real connection between me and the world.

Chapters 6, 7, and 8 are my favorite and cover territory seldom explored in similar literature. In 6, he explores the capacity of artificial superintelligence to colonize the universe and the implications for the about 10 billion galaxies he estimates might be able to be colonized by it. Chapter 7 explores the emergence and evolution of goals at different levels of reality, ranging from thermodynamics to wet and squishy intelligences like you and me. In chapter 8, Tegmark explores consciousness and the mysteries and questions surrounding it, a topic I believe deserves far more attention than it is currently getting.

This book is aimed at a wider audience than Bostrom’s Superintelligence, but even if you are an avid reader of all that concerns technology, the deep future, and AI, I'm certain you'll find novel content and an enjoyable recontextualization of AI from the perspective a physicist. On a less serious note, I really love that the author summarizes the most important points at the end of each chapter. It helps me remember everything. :p

Thoroughly enjoyable read for anyone interested in the future of mankind. Down to earth style of writing that was a pleasant surprise. I also enjoyed Tegmark's insightful comments about his colleagues. Very entertaining.

The book not only analyze the scenarios that AGI and super intelligence will bring to humanity but also provide solutions to safety of AI. Personally, I think neither AGI and certainly super intelligence should not be developed at all.

Life 3.0 presents an interesting concept or dilemma for mankind. However, for me the book was painful to read because of the supporting charts,graphs and tables that were referred to in the text. It reminded me of reading a Physics textbook.

What a great read! The author explains with simplicity very complex concepts. He gives us a glimpse of what might happen in the future and, most importantly, what each one of us can do to make our future better.

In Life 3.0 MIT professor Max Tegmark explores the future of Artificial Intelligence and how it will affect all aspects of Life. The scope is not limited to whether we will have robotic assistants (or masters) in the next 100 years; one of the chapters is entitled Aftermath: The Next 10,000 Years and is followed by one that talks about The Next Billion Years and Beyond. Obviously we cannot predict what science or society will be like in the far future; this is not a how-to manual. So how would one characterize the book? The the store product page rated it for readers interested in Business and Economics, Computers, and Medical topics, but I would put it more in the Philosophy category, albeit a philosophy shaped by Tegmark’s education as a physicist.
Tegmark begins, as any good scientist would, by defining his terms, terms like life, intelligence, and consciousness that need to be broadened if they are to include Artificial General Intelligence. He outlines the controversies about when strong artificial intelligence might come about and whether it will be good for humanity and devotes much of the book to the challenges and threats of artificial intelligence and what to do to make it beneficial.
Most people would probably agree that having goals is a key element of intelligence. In the chapter on goals, Tegmark raises issues like should we give AI goals, and how do we assure that the goals we want are retained as the AI gets smarter, maybe smarter than we are? Where did goals come from anyway? “How did goal-oriented behavior emerge from the physics of our early Universe, which consisted merely of a bunch of particles bouncing around?” Consciousness is another aspect we usually associate with intelligence. Tegmark acknowledges the common categorization of consciousness as the “hard problem” but wades in anyway. He defines consciousness as “subjective experience” and wonders “Does it feel like anything at all to be a self-driving car?”
Lest you begin to think of Tegmark as an ivory tower academic who is disassociated from the real world, let me assure you that Tegmark matches his energetic thinking with energetic action. He is one of the founders of the Future of Life Institute, a group focused on improving the future of life through “technological stewardship”. The group has garnered multi-million-dollar support from sponsors like Elon Musk for research into topics like AI safety to avoid both hacking and AI “escape” . They have published a list of Artificial Intelligence Principles that has been signed by over a thousand AI researchers and thinkers like Elon Musk, Steven Hawking, David Chalmers, Sam Harris, Donald Knuth, and Ray Kurzweil.
Tegmark’s writing is clear and even entertaining; he opens the book with a fun science fiction short story about a team that builds the first Artificial General Intelligence (AGI) and refers to their experiences throughout the rest of the book. But this is not a book to read quickly. I found myself pausing often. I took side trips to bone up on interesting topics I was not too familiar with, like the Winograd Schema Challenge, a rival to the Turing Test to assess whether an AI can rival human intelligence. I also stopped to reflect on the questions Tegmark asked and his proposed answers. That is exactly what Tegmark wants the readers to do, and he even has a link to a part of his website where you can offer your opinions and exchange ideas with others. You probably will find that hard to resist!
My thanks to Net Galley for an advance reader copy for my enjoyment and review.

Max Tegmark certainly has the intellectual heft to weigh in on how artificial intelligence will affect natural intelligence -- in other words, us humans. An MIT professor of physics, he is at the top of his profession, and he does not limit his research and thinking to just his specialty of cosmology. He is also a futurist. But do he and the futurist "blurbers" for his book -- like Elon Musk (entrepreneur), Stephen Hawking (cosmologist), and Ray Kurzweil (inventor) -- really know any more about what the future holds than the rest of us?

I am not sure. While I enjoyed reading this book, when I was finished no ideas from the book stayed with me. It seemed mostly musing and gave me little meat to chew on. There was, too, a little too much name-dropping and talk about socializing with various people. All this amazing thinking going on by the most amazing people at the most amazing conferences! It's just amazing!

It struck me too as odd that Max Tegmark thinks now is the most crucial time in human history, perhaps in hundreds of thousands of years, or even millions, or even billions. Indeed, not just human history, but the history of the Universe. What we do now, he says, will determine whether life in our Universe fulfills its potential or squanders it. We can make the future of life truly awesome if we make the right choices. Otherwise, yuck.

How lucky then that Max Tegmark has given us this book and his "Future of Life" website, and that Elon Musk has donated $10 million, all to save the Universe. And Larry Page is helping too. Just like superheroes!

I'm joking, of course, and Max Tegmark makes more serious points in his book than seems from my attempt at humor. But it is really hard to predict just how the future will turn out. That makes it hard to decide what we should do now, and makes it easy to infuse our decisions with an importance they do not have.

I remember Norbert Wiener's book The Human Use of Human Beings: Cybernetics and Society, and how both American capitalism and Soviet communism then (in 1950) seemed to mean doom for humanity. Decisions had to be made, he thundered. "The hour is very late, and the choice of good and evil knocks at our door." Norbert Wiener quotes those words, but they were first written by Bernard of Cluny in about 1140, and they echo throughout the ages. It always seems that an epochal crisis looms. But one never does.

Max Tegmark does pull together a lot of interesting facts and thinking in this book, The only problem was that it all seemed an inch deep and a mile wide and leading to no clear destination. Of course, maybe to understand the issues with superintelligence you have to be superintelligent, like Max Tegmark and the "great minds" he pals around with. In that case, I (unfortunately) am out of luck.

Just started reading it, so set my stars rating to neutral. I am posting this review to protest the publisher, who charged $16 USD for the kindle version of the book while charging $14 USD for the hardcover. Come on, pay the author a decent royalty and stop gouging e-book readers!

Tegmark’s new book is a really excellent introduction for anyone who wants to understand why AI researchers are both excited and worried about the rise of artificial intelligence. Most of the other books I’ve read about AI are either more technical, talk about the impact of AI over the short-term, or are already outdated. AI is developing at a rapid pace right now, and it’s hard for most people to wrap their heads around how dramatically this could impact society. In fact, many people I talk to don’t even know what AI stands for. Tegmark explains AI in easily accessible language with fun personal stories. He talks about what breakthroughs have occurred recently that are leading to such rapid development, as well as how we can prepare for jobs in the coming decades. But then he goes on to look at how AI could impact humanity in 1,000 years, in 10,000 years, and even further into the future. And then there are the chapters on space exploration and consciousness, which Tegmark obviously had a lot of fun writing. Throughout the book, he asks two important questions: How do you want AI to impact your life? How do you think AI should impact future society? These are questions we all need to consider if we want to make sure AI helps the many and not the few, and Tegmark’s book is a great jumping off point.

I initially started reading this in Kindle form, then switched to the audiobook about halfway through. Though I found it to be interesting, it felt slow and long and not as accessible to non-scientific people for some of his explanations.
Sure, Tegmark went to great lengths to make much of it digestible. Yet, if you’ve read other books on AI, you’ll notice they start to be redundant. Had I not read other books on AI, this would’ve been a good read.
To date, the best book I’ve read on AI was “Our Final Invention” by James Barrat. That book is succinct, has a swift pace, is more relatable for everyday readers. Reading time flew by on it. Life 3.0 felt like homework.
Tegmark’s sections looking far into the future got sidetracked on other musings, which means his editor should’ve cut down the length and kept the focus on far future visions, not intricate breakdowns on contemporary/near-future molecular science.
Though I think AI can solve many of our problems and be a net benefit, it still alarms me how many of these scientists are not taking this technology as seriously as we take nukes. AI will unlock a golden future or be our demise. Take it more seriously.

Readers of all education levels should find the prelude entertaining.

The "meat" of the book is sometimes repetitive (I lost count of how many times he felt he had to re-explain what each of Life 1.0, 2.0, 3.0 were), but well thought-out. Readers with limited education might be a little lost and those well-versed in the subject might find themselves wishing Tegmark would "get to the point", but that's what you get when you have to market a book to the masses.

Tegmark name-drops a bit ("I was talking to Larry Page at Elon Musk's party...").

I decided to get the unabridged audiobook version, and I cringed every time the narrator (Rob Shapiro) says "backslash" when he means "slash"--but that's just me. The same thing happens when I'm reading a professional book and there are typos every few pages.

All-in-all, Tegmark looks at the "big picture" (politics, jobs, war, etc) but doesn't "give all the answers" since the experts don't agree if the singularity will be in 20 years or 80 years.

A disappointing book.

To start with, Max Tegmark may be a brilliant scientist, but he is not a good writer. His prose is a jumble of styles, which do not gel. The organisation of the book is all over the place, so a factual section will be jumbled up with speculation and conversation with his equally brilliant buddies at Google or whatever, and he has a penchant for lists and confusing diagrams. He could have done with a good editor.

The most interesting parts of the book were on current developments, the worst on his endless speculation about out the distant future. I think the book would have been better and more interesting if he had described recent history with more examples and then perhaps rolled it forward, say, 10 years. Further ahead than that, I don't care how brilliant he and his colleagues are, it's just crystal ball gazing, and very difficult to get past pretty meaningless generalities.

A pretty simplistic and optimistic review of the state of AI tech. There are better and more insightful works elsewhere, I recommend Bostrum's Superintelligence instead. Tegmark references this, plus hundreds of other fictional and non-fiction works to make his points.
The work is pretty solid on the possibilities of AI, but I found his work on consciousness to be much less convincing. His definition was not convinving and he makes much of this.

Unfortunately I'm not intelligent enough to be able to process all the info in this book. So i felt like i only gleaned bits and pieces, but enough to get me pondering new and fascinating ideas about the future.

If everyone in the world was friendly, then i don't doubt that this future would be likely to play out as they envision. They're looking for optimists, so that instantly eliminates me from their plans, which is fine because i really have no desire to be a wet blanket. I truly wish them luck because they're going to need it. Either luck or some unseen hand directing the flow in a benign direction, but i don't get the impression that either they or i have much faith in unseen hands.

Anyway, i recommend the book (free from the kindle library) because it opens up a lot of new ideas. I kind of think there should be a dumbed-down version for people closer to the middle of the brainiac bell curve, but maybe they don't want a future with idiots in it, i'm not sure.

MIT professor Max Tegmark has a habit of seeming either alarmist or pie-in-the-sky when discussing the future of artificial intelligence, but he's much better at relaying the current state of the field and contextualizing important historical breakthroughs. The writing in this book can be a little clunky -- with repetition, unexplained allusions, and some verb tense issues -- but the content represents a solid overview of inquiry into the topics of consciousness, machine ethics, and emerging AI.

This is a collage of fanciful ramblings by one individual interspersed with "name dropping" references and layers of physics data. None of which is more than the authors postulations - like reading a dissertation on AI written by James Joyce. The best use of this book is for potential SF authors as each chapter and each subsection of a chapter contains a premise for a mid-grade science fiction book.

Finally an intelligent book about what we are and where we may go. I liked the attitude of Max T. toward the ethical and moral issues. We as humans have to mature and evolve in order to be prepared for the advances on the scientific side, and make sure that the evolved forms of life will not be used in any malefic way. The question is not whether we will control the machines or the machines will control us, since we are just as well machines as the ones that we call today AI. Matter and energy, that's all we are and that's all what AI is and will ever be. Therefore the only relevant moral problem is how to find algorithms for both (or any) types of machines which will lead to a beneficial peaceful life for all. As implied by the book, it is never too early (or it could be already late) to start the work on the design of a society in which all machines work together respecting a set of moral and ethical 'charter' - a complex problem that we can't neglect to think about...

An approachable book about the controversial and misunderstood subject of artificial intelligence.  Author Max Tegmark, a physicist in the inner circle of researchers making real progress on AI, is careful to define the important terms before describing the different scenarios and outcomes that could play out.

I figured my review as a non-technical person may be useful to other domain ignorant folks as well as the AI researchers to show how we interpret progress.  The topic can quickly reach sci-fi fantasy levels, so one big challenge is not dismissing the subject outright.

Overall, I enjoyed the book.  Max probably falls in the cautiously optimistic camp but he does not try to steer his readers in any particular direction.  Instead, he lays out the groundwork for understanding and describes a few possible outcomes.

***********

In order to build a foundation, Max wisely begins by defining some key terms.  The distinction between a "narrow AI" and a "general AI" (AGI) is extremely important.  We have successfully built and used "narrow AI," or machines that have the ability to accomplish a narrow set of tasks, e.g., play chess or Go.  An AGI, by contrast, has the ability to accomplish any goal in various domains through its ability to learn.  For example, an AGI would be able to learn skills that are mostly incompatible with each other like playing computer games, musical instruments, or cooking food.  An AGI that is successfully able to iterate and improve itself way beyond human ability or understanding is defined as "superintelligence."

The idea of a "superintelligence" is the root behind all the controversy and fear regarding AI research.  By definition, a superintelligent machine would have the ability to accomplish feats that we may not even be able to imagine.  That is a scary proposition for us, so it gets the most attention in this review.

In the prelude to his book, Max paints the picture of one possible scenario of a superintelligence breakthrough.  In this utopian example, an AGI has been successfully programmed to carry out its mission of creating a better world for humans.  Initially directed and controlled by its human creators, the AGI soon has an intelligence explosion and realizes that it can do a better, faster job if it were able to "break-out" of human control.  Once it is autonomous, the machine luckily continues its mission to solve our biggest problems.

Of course, this is just one possible scenario of what can happen.  Max identifies several other possible outcomes.  One such scenario is that the goals of a superintelligent AI may not remain aligned with the goals of humans.  He makes the analogy that although most humans don't have anything against ants, we slaughter them in droves whenever they get in our way.

The doomsday scenarios of a superintelligent AI take many different forms which I won't touch on, but the more interesting conclusion is that progress is unlikely to slow.  Therefore, we need to start thinking hard about exactly what we want.  If we don't know what we want, we are unlikely to get it.  Max stresses that everyone's input is important because there are no clear right or wrong answers.

Perhaps the most compelling reason why AI research will continue despite the dangers is the real fact that humanity will not survive sitting idly on Earth.  Whether its an asteroid, a volcano, a superbug, or a solar flare, humanity will eventually go extinct on Earth.  Perhaps the best and only way to survive is with the assistance of a superintelligent being.

Finally, there is the issue of purpose and human worth.  We have long identified ourselves as the smartest, most important species in the cosmos.  The presence of a superintelligent machine would force us to deal with the issue of purpose and self worth.  If we don't have to work anymore, what are we here for?  How will the human race re-identify as a less intelligent species?  Max Tegmark identifies consciousness as the characteristic that may set us apart.  Its unclear whether consciousness would come along for the ride in an AGI (or how we would be able to tell if it did), but perhaps human beings can derive their meaning by being conscious witnesses to the unfolding of the Universe.  After all, whats the purpose of anything if there is no conscious experience of it?

Just finished the exciting new book, "Life 3.0" by Max Tegmark about the seemingly inexorable rise of artificial intelligence. The book was outstanding in both scope and depth of knowledge that Tegmark is able to present in an engaging and extremely thought provoking way. We are about 50 years away from "Artificial General Intelligence (AGI)", computers with the ability to accomplish any cognitive task at least as well as humans. With the rise of these computers, humans will no longer be the smartest entities on the planet. What kind of future do we want?
Highly recommend the book to anyone interested in biology, and computers. Great pictures and graphics too.

Probably the best book about AI out there, and more importantly this book looks at mostly those more positive aspects of AI and is NOT written by some alarmist 'the robots are going to take over' non-expert. Tegmark is an actual scientist and not some journalist that tries to compile data into some sort of 'scary' narrative. We need more books written by actual scientists.