Scan barcode
tbauman's review
It's always interesting to read a book featuring people you know.
topperfalkon's review
4.0
I found this book to be an interesting but somewhat difficult read. Parts of the book felt quite informationally dense, making it harder to put down and dive back into.
Despite the title, the book only half focuses on AI safety, with much of the rest of the book covering the community that helped bring AI safety into the mainstream. That's not to say the topic shift is without merit, and for anyone unfamiliar with the Rationalist movement gives an interesting introduction. On the flip side, the introduction to the Rationalists is also a largely sympathetic take and although touches on some of the more controversial criticisms of the movement largely dismisses them.
If you're interested in AI or AI safety and its modern history, this book gives a good jumping off point.
Despite the title, the book only half focuses on AI safety, with much of the rest of the book covering the community that helped bring AI safety into the mainstream. That's not to say the topic shift is without merit, and for anyone unfamiliar with the Rationalist movement gives an interesting introduction. On the flip side, the introduction to the Rationalists is also a largely sympathetic take and although touches on some of the more controversial criticisms of the movement largely dismisses them.
If you're interested in AI or AI safety and its modern history, this book gives a good jumping off point.
hammo's review
4.0
A previous review mentioned that Gwern makes a cameo. TBH, this was what I was most excited about because he seems like a fascinating character who I'd like to learn more about. Unfortunately, he was only mentioned once, so I didn't learn anything about him. I already knew most of the content of this book, but here's the few things which were new to me:
* A disproportionate number of Tetlock's superforcasters came from Lesswrong
* Story about a rationalist woman who is considering having a baby, so starts carrying a robot baby around to get a sense of what the disruptions to her life would be like so she can make a more informed decision.
* Strategy for efficiently splitting bills: randomly generate a number from 0 to the total, then run a cumulative sum down the bill, and whoevers meal is on the random number has to pay for everything. Thus only one person pays and on average you pay for what you order.
* A disproportionate number of Tetlock's superforcasters came from Lesswrong
* Story about a rationalist woman who is considering having a baby, so starts carrying a robot baby around to get a sense of what the disruptions to her life would be like so she can make a more informed decision.
* Strategy for efficiently splitting bills: randomly generate a number from 0 to the total, then run a cumulative sum down the bill, and whoevers meal is on the random number has to pay for everything. Thus only one person pays and on average you pay for what you order.
baghaii's review
2.0
The problem with this book is that the title does not align with what the book is about very well. I had hoped that this book would be about AI, but it was more about the Rationalists and their concern about far future AI risk. It spent too much time on Less Wrong and Overcoming Bias and not enough time on actual AI.
One of the people that this book interviewed was Holden Karnofsky who is concerned with AI Risk and synthetic pandemics. I would imagine that Karnofsky would have a lot to say once we fell into a real pandemic. However, I am having trouble finding any work relevant to the current pandemic.
My problem with this book is that people concerned with far-future problems do not seem to be doing deep work that lends itself to near-future problems.
The author is overly generous with these folks because they have high-functioning autism. However, in real life, I do not see that spirit of generosity extended to autistic people who do not look like the people being interviewed for this book.
One of the people that this book interviewed was Holden Karnofsky who is concerned with AI Risk and synthetic pandemics. I would imagine that Karnofsky would have a lot to say once we fell into a real pandemic. However, I am having trouble finding any work relevant to the current pandemic.
My problem with this book is that people concerned with far-future problems do not seem to be doing deep work that lends itself to near-future problems.
The author is overly generous with these folks because they have high-functioning autism. However, in real life, I do not see that spirit of generosity extended to autistic people who do not look like the people being interviewed for this book.