Regular price: $17.00
Buy Now with 1 Credit
Buy Now for $17.00
Written by a computer scientist, this book was quite informative, and had a good grounding on how AI will develop in to a Super Intelligence, which is rarely discussed in such a prevailing subject.
It is quite unnerving to think the pursuit of a strong super AI is analogous to the development of weapons with the prisoners dilemma logic, making it an arms race of sorts. The author might be quite right to think Pandora's box has been opened, and there's no going back.
This book is somewhat unique in its acceptance that Super Artificial Intelligence will surpass humans at every level and become sentient, but optimistic in its predictions of a benevolent AI.
A good book with a positive attitude towards what's coming.
3 of 3 people found this review helpful
If a computer can ever just beat a human champion at checkers, then we’ll know they can think. No wait, better make that chess. No, no, we better make that the game of Go. Go needs intuition and only humans have intuition. Oh, never mind. Computers can never think or be self aware because that’s what most people say and the consensus just must always be right! For a mostly consensus defending awful book one can read ‘What do you think about machines that think’ available at Audible, but I really don’t recommend it.
Well this book lays out a contra road map of how it’s currently happening and speculates on why it will happen with certainty and how fool hardy it will be to deny it or try to outlaw it. The future is going to happen regardless of how one wants to control it.
The author is very good at telling stories. As he telling his stories, gems kept popping up. For example, when a doctor doesn’t know how to solve what ails you, they’ll end up doing one of three things: 1) give you a steroid to treat a possible inflammation, 2) give an antibiotic to kill an invader or 3) give a blood thinner like aspirin; another example, in the realm of war ‘amateurs talk about tactics, professionals talk about logistics’; my last example, in the realm of finance, ‘and that man was John Maynard Keynes’ (I’ll let you read the book to get the whole story, but it involves shorting pesetas).
The author questions our purpose. Why are we here? Is the universe a computer? He answers the first question by quoting Camus. I’ll refute Camus’ myth of Sisyphus by saying we have to learn and discover for ourselves. That is why I am here. Sure, one first needs to meet the daily needs and grind of existence, but after relaxation and needs are met learning is what motivates me.
AI (and ultimately AGI, artificial general intelligence) is one way we’re discovering and improving our world and discovering our meaning for being, and the author gives examples in the medical, security (computer and defense), financial, mind hacking, and other areas of life. Deep learning is real and is happening now. Computer software is teaching itself how to solve goal oriented actions.
The author had read 'Homo Deus' and knows that we create our own reality from the fictions we choose to believe in at the time and that is how AGI can ultimately develop too (our reality is a 'parler de facon').
Every experience we ever have is of a particular instant. We as humans try to make sense of our experiences by going from the specific to the general by using our intuition about the matter of fact. Four different times in different paradigms the author brought this concept up. Firstly (according to the author), Warren Buffett makes his money by feeling the quality of the management team of the company he wants to invest in and not just analytically analyzing financial reports and trends; secondly, he'll say the world is explained by the ‘structured and unstructured’ knowledge we have about the world; thirdly, the author uses Kahneman’s S1 and S2, quick logical thinking and slow feeling thinking. Lastly, in a story related by the author as told by his father the difference between Western thinking and Eastern thinking is analytical verse synthetic story telling. Overall, the author is defending the new age of deep learning ANI (artificial narrow intelligence) with an ability to feel (intuit) as well as logically (analytically) solve complex problems and soon the process will lead to the higher level AGI (artificial general intelligence).
Sometimes the author was out of his depth. Is Goldman Sachs profitable because of its proprietary trading as the author says or because they knew how to get ahead of the trade by a fraction of a second? Were the Russian Bots influential because a crazy racist uncle liked their articles on Facebook for their goofy conspiracy crap such as ‘climate change is a Chinese hoax’ as the author implies or was it because the reality based media printed nonsense stories about an email server almost every other day? There’s no doubt that reality based media has more influence then what the author called ‘mind hackers’ did. It turned out some of our gate keepers such as Matt Lauer, Charlie Rose, Mark Hallerpin all thought it was acceptable to ‘grab women inappropriately’ and they influenced public opinion and clearly saw women as mostly as objects and are thankfully no longer on my TV. Lastly, can anyone design an experiment such that S1 and S2 can be shown not to be true. If they can’t, then it’s pseudoscience. That doesn’t mean it’s not necessarily true, but it does mean that it lacks a foundation and should probably not be used as a foundation for other justified true beliefs.
Overall, the book is enjoyable. It doesn’t really break new ground or anything, but the author is a good story teller and this book is better than most AI books.
1 of 1 people found this review helpful