Other Sellers on Amazon
+ $3.99 shipping
AI Does Not Hate You Hardcover
Enter your mobile number or email address below and we'll send you a link to download the free Kindle App. Then you can start reading Kindle books on your smartphone, tablet, or computer - no Kindle device required.
To get the free app, enter your mobile phone number.
Frequently bought together
Customers who viewed this item also viewed
There was a problem filtering reviews right now. Please try again later.
Chivers works hard on being evenhanded, spending a fair amount of time talking to the subculture's fiercest critics and tackling those topics head-on. Those chapters were a bit awkward for me to get through, but I understand why they had to be there.
Overall, this is the first popular book or article that I can recommend to friends and family who want to understand my community, and I deeply appreciate that.
Overall, this is a book I recommend for anyone interested in a short treatment of the beliefs and history of the Rationalist Community (a movement of people who tried to have a deep think about the trajectory of human history and concluded we’re in far more peril than common sense would suggest). Like Chivers, I am familiar with the movement but not part of it, so I empathized with his approach to the topic: always sympathetic to his characters, but never quite welcomed in.
The book has several strengths and a few profound weaknesses. Here are a few in no particular order:
Strength: Chivers has a reasonable scope. He tries hard to focus in on one particular Rationalist concept: that an artificial intelligence that is smarter than a person would be a tremendously dangerous and destabilizing thing, and it may be coming sooner than we’re prepared to deal with. He could have written more deeply about Rationalist decision making or prediction methods, or about their views on charity, or about the various personalities under the Rationalist umbrella, but he always brings himself quickly back to the issue of AI safety. It’s a fast pace, touching on points just as needed to make sense of them, and then moving on. It’s almost too brisk (chapters are seldom more than five pages).
Weakness: Chivers has a weird lack of confidence in his own writing and in the rationalist ideas he’s presenting. Over and over, he describes a challenging argument or conclusion—something counter-intuitive or controversial—but instead of trusting the idea to have merit and his description to communicate why rationalists accept it (and why the reader should too), he retreats to a strange defensive crouch: “I know it sounds daft, but please believe me, this isn’t crazy, seriously, give it a chance.” He seems to expect the reader to approach the book with a dagger out, but frankly, if I am not convinced by an argument, pleading for me to reconsider probably won’t work.
Strength: Rationalists have assembled a nifty toolkit for increasing human effectiveness, and Chivers does a decent job repackaging some of those tools. This book doubles as a tight little self-help tome (“Thinking Like a Rationalist”)—perhaps a bit too Silicon Valley-flavored to be useful to many of us, but applicable enough that I feel as though I’m thinking a bit more clearly after finishing it. Of course, it’s possible that my brain has just tricked itself into thinking that it’s working better; I’m also even more skeptical of my own perceptions.
Weakness: Chivers implies he’s introducing us to a colorful cast of characters, but very little personality comes through. Part of this is due to brevity—with only a page or two on his topics, most of the thinkers and experts he interviews just don’t have the room to make an impression. But it’s more systematic than that: Chivers seems far more interested in the ideas of the Rationalists than he is in the people themselves. This is a defensible choice, but I can’t help but feel it leaves some cash on the table. One glaring example: He indicates, repeatedly, that Eliezer Yudkowsky (the grungy hobbit god-emperor of the Rationalist movement) is a wild and powerful personality: able to wow billionaires, woo impressionable ingénues, invent new schools of philosophy over a three-day weekend, blow up at the slightest provocation, etc. etc., but in the end we get very little sense of the man. Chivers was never able to meet him in person, and the physical absence of Rationalism’s central character is a real drag. It’s exacerbated (though perhaps entertainingly so) by the fact that Yudkowsky intermittently pops into the narrative to insult and belittle Chivers by email. I came away with the impression of a man who was simultaneously a titan and a toad, and whether this is Yudkowsky or Chivers’ fault, it’s a problem for the book.
Strength: The book seems fair. Very fair. Maybe too fair. Maybe obsessed, above all things, with being able to say that it was, no matter what else went wrong, FAIR. I think Chivers believes that Rationalism has revealed some important, scary things (I agree with this). And he wants to tell people about that. And he’s concerned that most people are going to react negatively, either because they don’t like to think about rogue AI or a bio-weapon apocalypse or whether they give enough to charity… or because they just don’t dig the geeky tech-crowd messengers. And he’s absolutely determined that if you blame those messengers, it won’t be because he didn’t go over and above, page-in and page-out, to be fair to them. He succeeds on this point—he tries to present a level-headed, balanced view of the movement and their beliefs. Good for him.
Weakness: I may be a bit spoiled having read Scott Alexander’s take on many of these topics (a rationalist blogger and a tremendously talented writer), but I found Chivers’ attempts at broaching complex topics hit-or-miss. His metaphors are adequate, his style is witty, his prose is flowy-as-you-please, but I don’t recall many cold-shivers of realization, or put-the-book-down-and-stare-into-the-horizon intermissions. The book is good, better than good, certainly better than I could do. But with the scope and magnitude of the universe-rattling ideas he recounts, the text is largely Earth-bound.
I rate it 4/5, on the strength of the seriousness with which Chivers approached his topic and the possibly-grave importance of the work. It’s a fast read, easily finished in one or two sittings, and is a fine introduction to an odd philosophical counterculture of the 21st Century. Heck, if the Rationalists are right, this book might be the singularly most important thing you ever read, by a dozen orders of magnitude. But it also skimps on color and character, sometimes misses on technical explanations, and is oddly inclined to apologize for itself. I hope it’s available in the US soon—the international shipping was a bear.
Top international reviews
I'd like to buy the author a beer, so if you're reading this, Tom, and you're ever anywhere near Chester, give me a shout.
If you're not Tom, just buy the book.
He seems to rely on people like Yudkowsky who is an AI expert but then spends most of his time talking about stuff totally unrelated to AI. Why didn't he speak to a lot of AI experts as well?
I am 75% the way through and have found it extremely disappointing. Not sure I will make it to the end which is extremely rare for me, I can only recall once before I have stopped reading a book before the end.
The successor to AlphaGo was never programmed 'how to learn or how to play' just these are the rules. it trained itself in a few days and now beats AlphaGo every time.
So can we manage AI at all?
Having delved in to this space via the Josh Clark Podcast "End of the World" I was glued to the book throughout. It's a riveting look at where we're inexorably heading as a species.