Worked for 26s
Here are the sources for the quotes on the page from Roman V. Yampolskiy’s chapter “Uncontrollability” (previously published as “On the Controllability of Artificial Intelligence: An Analysis of Limitations,” Journal of Cyber Security and Mobility, 2022).Researchgate
“One thing is for sure: We will not control it” — Elon Musk
From Elon Musk’s appearance on the Joe Rogan Experience podcast (circa 2018). In discussing digital superintelligence / the singularity, Musk said something to the effect that the outcome could be terrible or great, but “one thing is for sure: we will not control it.” This aligns with his broader comments in that episode about AI eventually being “outside of human control.” YouTube
“Dave: Open the pod bay doors, HAL. HAL: I’m sorry, Dave. I’m afraid I can’t do that” — HAL 2001
Exact dialogue from the 1968 film 2001: A Space Odyssey (directed by Stanley Kubrick, screenplay by Kubrick and Arthur C. Clarke). Dave Bowman (Keir Dullea) requests access; the HAL 9000 computer (voiced by Douglas Rain) refuses. This is one of the most famous lines in cinema (AFI top 100 movie quotes). Moviequotedb
“[F]inding a solution to the AI control problem is an important task; in Bostrom’s sonorous words, ‘the essential task of our age’” — Stuart Russell
From Stuart Russell’s essay “If We Succeed” (Dædalus, 2022, American Academy of Arts & Sciences). Russell writes: “Finding a solution to the AI control problem is an important task; it may be, in the words of philosopher Nick Bostrom, ‘the essential task of our age.’” The Bostrom phrasing comes from Superintelligence (2014) and related writings on the control problem. Amacad
“[I]t seems probable that once the machine thinking method had started, it would not take long to outstrip our feeble powers. … At some stage therefore we should have to expect the machines to take control” — Alan Turing
From Alan Turing’s 1951 lecture/paper “Intelligent Machinery, A Heretical Theory” (delivered to the ’51 Society, Manchester). Full relevant passage: “...it seems probable that once the machine thinking method had started, it would not take long to outstrip our feeble powers. There would be no question of the machines dying, and they would be able to converse with each other to sharpen their wits. At some stage therefore we should have to expect the machines to take control, in the way that is mentioned in Samuel Butler’s Erewhon.” Uberty
“Prohibit the development of Artificial Intelligence capable of saying ‘no’ to humans” — 赵汀阳 (Zhao Tingyang)
From Chinese philosopher Zhao Tingyang’s writings on AI ethics/ontology, particularly his 2018 essay “Near-term Worries and Long-term Concerns of the Artificial Intelligence ‘Revolution’” (and related pieces such as “If Artificial Intelligence Says ‘No’ to Humans”). He argues that the key safety principle is to prohibit development of AI capable of refusing human commands / saying “no” to humans, as that capacity would mark a civilizational rupture comparable to humans first gaining the ability to refuse. Kunlunce
“Any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes” — Charles Goodhart
The original formulation of Goodhart’s Law, from British economist Charles Goodhart’s 1975 paper/presentation on monetary policy (later published in Monetary Theory and Practice, 1984). The more common popular paraphrase is “When a measure becomes a target, it ceases to be a good measure.” Wikipedia
The page itself is the opening of Chapter 6 from Yampolskiy’s work on AI uncontrollability (also available as an arXiv preprint and journal article).
Here are the sources for the quotes on the page from Roman V. Yampolskiy’s chapter “Uncontrollability” (previously published as “On the Controllability of Artificial Intelligence: An Analysis of Limitations,” Journal of Cyber Security and Mobility, 2022).Researchgate
“One thing is for sure: We will not control it” — Elon Musk
From Elon Musk’s appearance on the Joe Rogan Experience podcast (circa 2018). In discussing digital superintelligence / the singularity, Musk said something to the effect that the outcome could be terrible or great, but “one thing is for sure: we will not control it.” This aligns with his broader comments in that episode about AI eventually being “outside of human control.” YouTube
“Dave: Open the pod bay doors, HAL. HAL: I’m sorry, Dave. I’m afraid I can’t do that” — HAL 2001
Exact dialogue from the 1968 film 2001: A Space Odyssey (directed by Stanley Kubrick, screenplay by Kubrick and Arthur C. Clarke). Dave Bowman (Keir Dullea) requests access; the HAL 9000 computer (voiced by Douglas Rain) refuses. This is one of the most famous lines in cinema (AFI top 100 movie quotes). Moviequotedb
“[F]inding a solution to the AI control problem is an important task; in Bostrom’s sonorous words, ‘the essential task of our age’” — Stuart Russell
From Stuart Russell’s essay “If We Succeed” (Dædalus, 2022, American Academy of Arts & Sciences). Russell writes: “Finding a solution to the AI control problem is an important task; it may be, in the words of philosopher Nick Bostrom, ‘the essential task of our age.’” The Bostrom phrasing comes from Superintelligence (2014) and related writings on the control problem. Amacad
“[I]t seems probable that once the machine thinking method had started, it would not take long to outstrip our feeble powers. … At some stage therefore we should have to expect the machines to take control” — Alan Turing
From Alan Turing’s 1951 lecture/paper “Intelligent Machinery, A Heretical Theory” (delivered to the ’51 Society, Manchester). Full relevant passage: “...it seems probable that once the machine thinking method had started, it would not take long to outstrip our feeble powers. There would be no question of the machines dying, and they would be able to converse with each other to sharpen their wits. At some stage therefore we should have to expect the machines to take control, in the way that is mentioned in Samuel Butler’s Erewhon.” Uberty
“Prohibit the development of Artificial Intelligence capable of saying ‘no’ to humans” — 赵汀阳 (Zhao Tingyang)
From Chinese philosopher Zhao Tingyang’s writings on AI ethics/ontology, particularly his 2018 essay “Near-term Worries and Long-term Concerns of the Artificial Intelligence ‘Revolution’” (and related pieces such as “If Artificial Intelligence Says ‘No’ to Humans”). He argues that the key safety principle is to prohibit development of AI capable of refusing human commands / saying “no” to humans, as that capacity would mark a civilizational rupture comparable to humans first gaining the ability to refuse. Kunlunce
“Any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes” — Charles Goodhart
The original formulation of Goodhart’s Law, from British economist Charles Goodhart’s 1975 paper/presentation on monetary policy (later published in Monetary Theory and Practice, 1984). The more common popular paraphrase is “When a measure becomes a target, it ceases to be a good measure.” Wikipedia
The page itself is the opening of Chapter 6 from Yampolskiy’s work on AI uncontrollability (also available as an arXiv preprint and journal article).
Disiderata:
105 sources
No. 5 indicates that you have corrected yourself, the added info allowing you to be more precise
Worked for 14s
Yes — the additional search results allowed a more precise sourcing for #5.
Refined source for quote 5:
“Prohibit the development of Artificial Intelligence capable of saying ‘no’ to humans”
— 赵汀阳 (Zhao Tingyang)
This is a direct English rendering of Zhao Tingyang’s core proposed safety principle from his 2018 essay “人工智能‘革命’的‘近忧’和‘远虑’——一种伦理学和存在论的分析” (“Near-term Worries and Long-term Concerns of the Artificial Intelligence ‘Revolution’: An Analysis of Ethics and Ontology”).
In the original Chinese he states (approximately):
“如果允许给出一个并且仅仅一个，那么我愿意说，只需要一个原则，即禁止研发有能力对人类说‘不’的人工智能。”
(“If only one principle is allowed, then I would say there needs to be only one: prohibit the development of artificial intelligence that is capable of saying ‘no’ to humans.”)
He elaborates that the ability of AI to refuse human commands would be a civilizational rupture comparable to the moment humans first gained the capacity to say “no,” marking the start of a new historical era no longer centered on humanity.
This matches the wording and attribution on the page exactly.
No comments:
Post a Comment