![[Review] Human Compatible: Artificial Intelligence and the Problem of Control (Stuart Russell) Summarized.](https://episodes.castos.com/660078c6833215-59505987/images/2636740/c1a-085k3-dmdqdxwpckwg-lrfwog.jpg)
Show Notes
Human Compatible: Artificial Intelligence and the Problem of Control (Stuart Russell)
- Amazon USA Store: https://www.amazon.com/dp/0525558632?tag=9natree-20
- Amazon Worldwide Store: https://global.buys.trade/Human-Compatible%3A-Artificial-Intelligence-and-the-Problem-of-Control-Stuart-Russell.html
- Apple Books: https://books.apple.com/us/audiobook/human-compatible-artificial-intelligence-and-the/id1479933486?itsct=books_box_link&itscg=30200&ls=1&at=1001l3bAw&ct=9natree
- eBay: https://www.ebay.com/sch/i.html?_nkw=Human+Compatible+Artificial+Intelligence+and+the+Problem+of+Control+Stuart+Russell+&mkcid=1&mkrid=711-53200-19255-0&siteid=0&campid=5339060787&customid=9natree&toolid=10001&mkevt=1
- Read more: https://english.9natree.com/read/0525558632/
#AIcontrolproblem #Humanpreferenceuncertainty #Inversereinforcementlearning #Beneficialmachines #Lethalautonomousweapons #HumanCompatible
Human Compatible: Artificial Intelligence and the Problem of Control is a 2019 nonfiction book by Stuart Russell, a computer scientist and prominent AI researcher. Written for general readers as well as technical and policy audiences, it examines a central challenge of advanced AI: how people can retain meaningful control over systems that may eventually exceed human capabilities in important domains. Russell distinguishes this problem from the familiar image of malicious machines. His concern is that an extremely capable system can cause grave harm while competently pursuing an objective that humans specified badly. The book combines an accessible history of AI, discussion of current and prospective applications, and an argument for changing the basic model of machine intelligence. Rather than designing systems that optimize fixed goals with ever greater power, Russell proposes machines that remain uncertain about human preferences and deferential to human input. The result is both a warning about AI misuse and long-term loss of control and a technically grounded case for beneficial AI research.
- Amazon USA Store: https://www.amazon.com/dp/0525558632?tag=9natree-20
- Amazon Worldwide Store: https://global.buys.trade/Human-Compatible%3A-Artificial-Intelligence-and-the-Problem-of-Control-Stuart-Russell.html
- Apple Books: https://books.apple.com/us/audiobook/human-compatible-artificial-intelligence-and-the/id1479933486?itsct=books_box_link&itscg=30200&ls=1&at=1001l3bAw&ct=9natree
- eBay: https://www.ebay.com/sch/i.html?_nkw=Human+Compatible+Artificial+Intelligence+and+the+Problem+of+Control+Stuart+Russell+&mkcid=1&mkrid=711-53200-19255-0&siteid=0&campid=5339060787&customid=9natree&toolid=10001&mkevt=1
- Read more: https://english.9natree.com/read/0525558632/
#AIcontrolproblem #Humanpreferenceuncertainty #Inversereinforcementlearning #Beneficialmachines #Lethalautonomousweapons #HumanCompatible
Human Compatible: Artificial Intelligence and the Problem of Control is a 2019 nonfiction book by Stuart Russell, a computer scientist and prominent AI researcher. Written for general readers as well as technical and policy audiences, it examines a central challenge of advanced AI: how people can retain meaningful control over systems that may eventually exceed human capabilities in important domains. Russell distinguishes this problem from the familiar image of malicious machines. His concern is that an extremely capable system can cause grave harm while competently pursuing an objective that humans specified badly. The book combines an accessible history of AI, discussion of current and prospective applications, and an argument for changing the basic model of machine intelligence. Rather than designing systems that optimize fixed goals with ever greater power, Russell proposes machines that remain uncertain about human preferences and deferential to human input. The result is both a warning about AI misuse and long-term loss of control and a technically grounded case for beneficial AI research.
