AI deception: A survey of examples, risks, and potential solutions
"Large language models and other AI systems have already learned, from their training, the ability to deceive via techniques such as manipulation, sycophancy, and cheating the safety test. AI’s increasing capabilities at deception pose serious risks, ranging from short-term risks, such as fraud and election tampering, to long-term risks, such as losing control of AI systems."
AI deception: A survey of examples, risks, and potential solutions
"Large language models and other AI systems have already learned, from their training, the ability to deceive via techniques such as manipulation, sycophancy, and cheating the safety test. AI’s increasing capabilities at deception pose serious risks, ranging from short-term risks, such as fraud and election tampering, to long-term risks, such as losing control of AI systems."
AI deception: A survey of examples, risks, and potential solutions
“Large language models and other AI systems have already learned, from their training, the ability to deceive via techniques such as manipulation, sycophancy, and cheating the safety test. AI’s increasing capabilities at deception pose serious risks, ranging from short-term risks, such as fraud and election tampering, to long-term risks, such as losing control of AI systems.”
AI deception: A survey of examples, risks, and potential solutions
“Large language models and other AI systems have already learned, from their training, the ability to deceive via techniques such as manipulation, sycophancy, and cheating the safety test. AI’s increasing capabilities at deception pose serious risks, ranging from short-term risks, such as fraud and election tampering, to long-term risks, such as losing control of AI systems.”
AI deception: A survey of examples, risks, and potential solutions
"Large language models and other AI systems have already learned, from their training, the ability to deceive via techniques such as manipulation, sycophancy, and cheating the safety test. AI’s increasing capabilities at deception pose serious risks, ranging from short-term risks, such as fraud and election tampering, to long-term risks, such as losing control of AI systems."
AI deception: A survey of examples, risks, and potential solutions
"Large language models and other AI systems have already learned, from their training, the ability to deceive via techniques such as manipulation, sycophancy, and cheating the safety test. AI’s increasing capabilities at deception pose serious risks, ranging from short-term risks, such as fraud and election tampering, to long-term risks, such as losing control of AI systems."
Classic Krakauer Essays on Wilderness and Risk by Jon Krakauer
Spanning an extraordinary range of subjects and locations, these articles take us from a horrifying avalanche on Mt. Everest to a volcano poised to obliterate a big chunk of greater Seattle at any moment; from a wilderness teen-therapy program run by apparent sadists to an otherwordly cave in New Mexico, studied by NASA to better understand Mars...