通用人工智能

来自集智百科
跳到导航 跳到搜索

此词条由袁一博翻译,未经人工整理和审校,带来阅读不便,请见谅。

本词条已由Ricky审校。


Artificial general intelligence (AGI) is the hypothetical[1] intelligence of a machine that has the capacity to understand or learn any intellectual task that a human being can. It is a primary goal of some artificial intelligence research and a common topic in science fiction and futures studies. AGI can also be referred to as strong AI,[2][3]

Artificial general intelligence (AGI) is the hypothetical intelligence of a machine that has the capacity to understand or learn any intellectual task that a human being can. It is a primary goal of some artificial intelligence research and a common topic in science fiction and futures studies. AGI can also be referred to as strong AI.

通用人工智能(Artificial general intelligence,AGI)是一种假想中的机器智能[4],它有能力理解或学习任何人类能够完成的智力任务。这是一些人工智能研究的主要目标,也是科幻小说和未来研究的共同话题。通用人工智能也可以被称为强人工智能(Strong AI),完全人工智能(Full AI),或通用智能行为


Some academic sources reserve the term "strong AI" for machines that can experience consciousness. Today's AI is speculated to be many years, if not decades, away from AGI.

一些学术文献保留了“强人工智能”这个术语,用来形容能够体会意识的机器。据推测,今天的人工智能将在很多年,甚至很多个十年之后才能达到通用人工智能的地步。


Some authorities emphasize a distinction between strong AI and applied AI,[5] also called narrow AI[3] or weak AI.[6] In contrast to strong AI, weak AI is not intended to perform human cognitive abilities. Rather, weak AI is limited to the use of software to study or accomplish specific problem solving or reasoning tasks.

Some authorities emphasize a distinction between strong AI and applied AI, also called narrow AI In contrast to strong AI, weak AI is not intended to perform human cognitive abilities. Rather, weak AI is limited to the use of software to study or accomplish specific problem solving or reasoning tasks.

一些权威机构强调强人工智能应用人工智能,或者说'狭义人工智能(weak AI)'强人工智能之间的区别:弱人工智能并不需要动用人类的认知能力。相反,弱人工智能仅限于用在通过软件来研究或解决特定问题,或完成推理任务。


As of 2017, over forty organizations are researching AGI.[7]

As of 2017, over forty organizations are researching AGI.

截止到2017年,已经有超过四十家机构在研究 AGI。


判定要求


Various criteria for intelligence have been proposed (most famously the Turing test) but to date, there is no definition that satisfies everyone.[8] However, there is wide agreement among artificial intelligence researchers that intelligence is required to do the following:

人们提出了各种各样的智能标准(最著名的是图灵测试(Turing test) ,但到目前为止,还没有一个定义能使所有人满意。然而,人工智能研究人员普遍认为,智能需要做到以下几点: [9]


  • 推理、使用策略,解决问题,并且在不确定条件下做出决策;
  • 表示知识,包括常识;
  • 规划;
  • 学习;
  • 使用自然语言交流;
  • 以及综合运用所有技巧以达到某个目的。


Other important capabilities include the ability to sense (e.g. see) and the ability to act (e.g. move and manipulate objects) in the world where intelligent behaviour is to be observed.[10] This would include an ability to detect and respond to hazard.[11] Many interdisciplinary approaches to intelligence (e.g. cognitive science, computational intelligence and decision making) tend to emphasise the need to consider additional traits such as imagination (taken as the ability to form mental images and concepts that were not programmed in)[12] and autonomy.[13]

Other important capabilities include the ability to sense (e.g. see) and the ability to act (e.g. move and manipulate objects) in the world where intelligent behaviour is to be observed. This would include an ability to detect and respond to hazard. Many interdisciplinary approaches to intelligence (e.g. cognitive science, computational intelligence and decision making) tend to emphasise the need to consider additional traits such as imagination (taken as the ability to form mental images and concepts that were not programmed in) and autonomy.

其他重要的能力包括在客观世界进行感知(例如:视觉)和行动(例如:移动和操纵物体)的能力。智能行为在客观世界中是可观测的。这将包括检测和应对危险的能力。许多跨学科的智力研究方法(例如:认知科学、计算智能和决策)倾向于强调考虑额外特征的必要性,例如想象力(被认为是未编入程序的形成意象和概念的能力)和自主性。

Computer based systems that exhibit many of these capabilities do exist (e.g. see computational creativity, automated reasoning, decision support system, robot, evolutionary computation, intelligent agent), but not yet at human levels.

Computer based systems that exhibit many of these capabilities do exist (e.g. see computational creativity, automated reasoning, decision support system, robot, evolutionary computation, intelligent agent), but not yet at human levels.

基于计算机的智能系统,展示出了许多上述能力(例如:计算创造性、自动推理、决策支持系统、机器人、进化计算、智能代理) ,但还没有达到人类的水平。


判定人类水平通用人工智能的测试

The following tests to confirm human-level AGI have been considered:[14][15]

The following tests to confirm human-level AGI have been considered:

经过了许多思考,下列测试被提出以确认机器是否拥有人类水平的通用智能:

图灵测试
A machine and a human both converse sight unseen with a second human, who must evaluate which of the two is the machine, which passes the test if it can fool the evaluator a significant fraction of the time. Note: Turing does not prescribe what should qualify as intelligence, only that knowing that it is a machine should disqualify it.
一个机器人和一个人类都与另一个人类在看不到的情况下进行交流,后者必须评估两者中哪一个是机器,如果它能在很大一部分时间内骗过评估者,那么机器就通过了测试。注意: 图灵并没有规定什么是智能,只规定了只要一台机器能被指认出来,那就不应该认为它是智能的。
The Coffee Test (Wozniak)
咖啡测试(沃兹尼亚克)
A machine is required to enter an average American home and figure out how to make coffee: find the coffee machine, find the coffee, add water, find a mug, and brew the coffee by pushing the proper buttons.
一台机器需要进入一个普通的美国家庭,并弄清楚如何制作咖啡: 找到咖啡机,找到咖啡,加水,找到一个马克杯,并通过按下正确的按钮来煮咖啡。
The Robot College Student Test (Goertzel)
机器人大学生考试(格兹尔)
A machine enrolls in a university, taking and passing the same classes that humans would, and obtaining a degree.
一台机器进入一所大学,学习并通过与人类相同的课程,并获得学位。
The Employment Test (Nilsson)
就业测试(尼尔森)
A machine works an economically important job, performing at least as well as humans in the same job.
机器从事一项经济上重要的工作,在同一项工作中表现得至少和人类一样好。


需要通用人工智能解决的问题

The most difficult problems for computers are informally known as "AI-complete" or "AI-hard", implying that solving them is equivalent to the general aptitude of human intelligence, or strong AI, beyond the capabilities of a purpose-specific algorithm.[16]

The most difficult problems for computers are informally known as "AI-complete" or "AI-hard", implying that solving them is equivalent to the general aptitude of human intelligence, or strong AI, beyond the capabilities of a purpose-specific algorithm.

对于计算机来说,最困难的问题被非正式地称为“AI完全问题(AI-complete)”或“AI困难问题(AI-hard)”,这意味着解决这些问题相当于拥有人类智能的一般才能,或超出了特定目的算法能力的强人工智能。


AI-complete problems are hypothesised to include general computer vision, natural language understanding, and dealing with unexpected circumstances while solving any real world problem.[17]

AI-complete problems are hypothesised to include general computer vision, natural language understanding, and dealing with unexpected circumstances while solving any real world problem.

AI-完全问题被认为包括一般的计算机视觉,自然语言理解,以及在解决任何现实世界问题的同时处理意外情况。


AI-complete problems cannot be solved with current computer technology alone, and also require human computation. This property could be useful, for example, to test for the presence of humans, as CAPTCHAs aim to do; and for computer security to repel brute-force attacks.[18][19]

AI-complete problems cannot be solved with current computer technology alone, and also require human computation. This property could be useful, for example, to test for the presence of humans, as CAPTCHAs aim to do; and for computer security to repel brute-force attacks.

目前的计算机技术不能单独解决AI完全问题,还需要人工计算的参与。这个特性可以用来测试人类是否存在(比如说,CAPTCHAs的目标就是测试服务的使用者是人类而非机器人) ,以及应用于计算机安全以抵御暴力攻击。

历史

经典人工智能

Modern AI research began in the mid 1950s.[20] The first generation of AI researchers were convinced that artificial general intelligence was possible and that it would exist in just a few decades. AI pioneer Herbert A. Simon wrote in 1965: "machines will be capable, within twenty years, of doing any work a man can do."[21] Their predictions were the inspiration for Stanley Kubrick and Arthur C. Clarke's character HAL 9000, who embodied what AI researchers believed they could create by the year 2001. AI pioneer Marvin Minsky was a consultant[22] on the project of making HAL 9000 as realistic as possible according to the consensus predictions of the time; Crevier quotes him as having said on the subject in 1967, "Within a generation ... the problem of creating 'artificial intelligence' will substantially be solved,"[23] although Minsky states that he was misquoted.[citation needed]

Modern AI research began in the mid 1950s. The first generation of AI researchers were convinced that artificial general intelligence was possible and that it would exist in just a few decades. AI pioneer Herbert A. Simon wrote in 1965: "machines will be capable, within twenty years, of doing any work a man can do." Their predictions were the inspiration for Stanley Kubrick and Arthur C. Clarke's character HAL 9000, who embodied what AI researchers believed they could create by the year 2001. AI pioneer Marvin Minsky was a consultant on the project of making HAL 9000 as realistic as possible according to the consensus prediction of the time; Crevier quotes him as having said on the subject in 1967, "Within a generation ... the problem of creating 'artificial intelligence' will substantially be solved," although Minsky states that he was misquoted.

现代人工智能研究始于20世纪50年代中期。第一代人工智能研究人员确信,通用人工智能是可能的,并将在短短几十年内出现。人工智能的先驱赫伯特·A·西蒙(Herbert A. Simon)在1965年写道: “机器将在20年内拥有完成人类能做的任何工作的能力。”他们的预言启发了斯坦利·库布里克和亚瑟·查理斯·克拉克塑造的角色“哈尔9000”,它代表了人工智能研究人员相信他们到2001年时能够创造出的东西。人工智能先驱马文·明斯基(Marvin Minsky)当时是一个项目顾问,该项目旨在根据当时的一致预测,使“哈尔9000”尽可能逼真; 克里维尔援引他在1967年关于这个问题的话说,“在一代人的时间里... ... 创造‘人工智能’的问题将大体上得到解决,”尽管明斯基声称,他的话被错误引用了。


However, in the early 1970s, it became obvious that researchers had grossly underestimated the difficulty of the project. Funding agencies became skeptical of AGI and put researchers under increasing pressure to produce useful "applied AI".[24] As the 1980s began, Japan's Fifth Generation Computer Project revived interest in AGI, setting out a ten-year timeline that included AGI goals like "carry on a casual conversation".[25] In response to this and the success of expert systems, both industry and government pumped money back into the field.[26] However, confidence in AI spectacularly collapsed in the late 1980s, and the goals of the Fifth Generation Computer Project were never fulfilled.[27] For the second time in 20 years, AI researchers who had predicted the imminent achievement of AGI had been shown to be fundamentally mistaken. By the 1990s, AI researchers had gained a reputation for making vain promises. They became reluctant to make predictions at all[28] and to avoid any mention of "human level" artificial intelligence for fear of being labeled "wild-eyed dreamer[s]."[29]

However, in the early 1970s, it became obvious that researchers had grossly underestimated the difficulty of the project. Funding agencies became skeptical of AGI and put researchers under increasing pressure to produce useful "applied AI". As the 1980s began, Japan's Fifth Generation Computer Project revived interest in AGI, setting out a ten-year timeline that included AGI goals like "carry on a casual conversation". In response to this and the success of expert systems, both industry and government pumped money back into the field. However, confidence in AI spectacularly collapsed in the late 1980s, and the goals of the Fifth Generation Computer Project were never fulfilled. For the second time in 20 years, AI researchers who had predicted the imminent achievement of AGI had been shown to be fundamentally mistaken. By the 1990s, AI researchers had gained a reputation for making vain promises. They became reluctant to make predictions at all and to avoid any mention of "human level" artificial intelligence for fear of being labeled "wild-eyed dreamer[s]."

然而,在20世纪70年代初,很明显,研究人员严重低估了该项目的难度。资助机构开始对通用人工智能持怀疑态度,并对研究人员施加越来越大的压力,要求他们生产出有用的“应用人工智能”。在20世纪80年代的初期,日本的第五代计算机项目(Fifth Generation Computer Project)重新唤起了人们对通用人工智能的兴趣,并设定了一个长达10年的时间线,其中包括通用人工智能的目标,比如“进行一次随意的交谈”。为了应对这种情况,以及专家系统的成功实现,工业界和政府都重新将资金投入这一领域。然而,人们对人工智能的信心在20世纪80年代末大幅下降,第五代计算机项目的目标从未实现。20年来的第二次,人工智能研究员预测通用人工智能即将取得的成果被证明根本是错误的。到了20世纪90年代,人工智能研究人员因做出虚假承诺而臭名昭著。他们变得彻底不再愿意做预测,也不愿意提及“人类水平”的人工智能,因为他们害怕被贴上“狂热梦想家”的标签


狭义人工智能的研究

In the 1990s and early 21st century, mainstream AI achieved far greater commercial success and academic respectability by focusing on specific sub-problems where they can produce verifiable results and commercial applications, such as artificial neural networks and statistical machine learning.[30] These "applied AI" systems are now used extensively throughout the technology industry, and research in this vein is very heavily funded in both academia and industry. Currently, development on this field is considered an emerging trend, and a mature stage is expected to happen in more than 10 years.[31]

In the 1990s and early 21st century, mainstream AI achieved far greater commercial success and academic respectability by focusing on specific sub-problems where they can produce verifiable results and commercial applications, such as artificial neural networks and statistical machine learning. These "applied AI" systems are now used extensively throughout the technology industry, and research in this vein is very heavily funded in both academia and industry. Currently, development on this field is considered an emerging trend, and a mature stage is expected to happen in more than 10 years.

在20世纪90年代和21世纪初,主流人工智能取得了更大的商业成功和学术声望,因为它们把重点放在能够产出可验证结果和商业应用的具体子问题上,例如人工神经网络和统计机器学习。这些“应用人工智能”系统现在在整个技术产业中得到广泛应用,这方面的研究得到了学术界和产业界的大量资助。目前,这一领域的发展被认为是一个新兴的趋势,并有望在10多年内进入一个成熟的阶段。


Most mainstream AI researchers hope that strong AI can be developed by combining the programs that solve various sub-problems. Hans Moravec wrote in 1988:

"I am confident that this bottom-up route to artificial intelligence will one day meet the traditional top-down route more than half way, ready to provide the real world competence and the commonsense knowledge that has been so frustratingly elusive in reasoning programs. Fully intelligent machines will result when the metaphorical golden spike is driven uniting the two efforts."[32]

Most mainstream AI researchers hope that strong AI can be developed by combining the programs that solve various sub-problems. Hans Moravec wrote in 1988:

"I am confident that this bottom-up route to artificial intelligence will one day meet the traditional top-down route more than half way, ready to provide the real world competence and the commonsense knowledge that has been so frustratingly elusive in reasoning programs. Fully intelligent machines will result when the metaphorical golden spike is driven uniting the two efforts."

大多数主流人工智能研究人员希望,通过把能解决各种子问题的程序结合起来,可以开发出强人工智能。汉斯·莫拉维克(Hans Moravec)在1988年写道: “我相信,这种自下而上的人工智能路线,终有一天会与传统的自上而下的路线在后半程相遇,然后提供能解决真实世界中问题的能力,以及常识知识——在推理程序中一直都难以捉摸的令人沮丧的东西。这两种路线结合的人工智能将能为我们解决这些疑难。当以后有一种神奇的方法把这二者结合起来时,完全智能的机器就会产生。”


However, even this fundamental philosophy has been disputed; for example, Stevan Harnad of Princeton concluded his 1990 paper on the Symbol Grounding Hypothesis by stating:

"The expectation has often been voiced that "top-down" (symbolic) approaches to modeling cognition will somehow meet "bottom-up" (sensory) approaches somewhere in between. If the grounding considerations in this paper are valid, then this expectation is hopelessly modular and there is really only one viable route from sense to symbols: from the ground up. A free-floating symbolic level like the software level of a computer will never be reached by this route (or vice versa) – nor is it clear why we should even try to reach such a level, since it looks as if getting there would just amount to uprooting our symbols from their intrinsic meanings (thereby merely reducing ourselves to the functional equivalent of a programmable computer)."[33]

However, even this fundamental philosophy has been disputed; for example, Stevan Harnad of Princeton concluded his 1990 paper on the Symbol Grounding Hypothesis by stating:

"The expectation has often been voiced that "top-down" (symbolic) approaches to modeling cognition will somehow meet "bottom-up" (sensory) approaches somewhere in between. If the grounding considerations in this paper are valid, then this expectation is hopelessly modular and there is really only one viable route from sense to symbols: from the ground up. A free-floating symbolic level like the software level of a computer will never be reached by this route (or vice versa) – nor is it clear why we should even try to reach such a level, since it looks as if getting there would just amount to uprooting our symbols from their intrinsic meanings (thereby merely reducing ourselves to the functional equivalent of a programmable computer)."

然而,连如此基本的哲学问题也存在争议; 例如,普林斯顿大学的斯蒂文·哈纳德(Stevan Harnad)在1990年关于符号基础假说(the Symbol Grounding Hypothesis)的论文中总结道: “人们经常提出这样的期望,即建立“自上而下”(符号)的认知模型的方法将在某种程度上与“自下而上”(感官)的方法在建模过程中的某处相会。如果本文中的基本假设是正确的,那就不可能存在模块化的后见方式,且从认知到符号真的只有一条可行的路径: 直接建立从感觉到符号的联系。类似计算机软件级别的无意义的符号永远不可能通过这条路径实现,反之亦然——甚至我们也不清楚为什么应该尝试达到这样一个级别,因为它看起来就像是把我们的符号从它们的内在意义上连根拔起(从而仅仅把我们自己变成和可编程计算机一样的东西)。”


现代通用人工智能的研究

The term "artificial general intelligence" was used as early as 1997, by Mark Gubrud[34] in a discussion of the implications of fully automated military production and operations. The term was re-introduced and popularized by Shane Legg and Ben Goertzel around 2002.[35] The research objective is much older, for example Doug Lenat's Cyc project (that began in 1984), and Allen Newell's Soar project are regarded as within the scope of AGI. AGI research activity in 2006 was described by Pei Wang and Ben Goertzel[36] as "producing publications and preliminary results". The first summer school in AGI was organized in Xiamen, China in 2009[37] by the Xiamen university's Artificial Brain Laboratory and OpenCog. The first university course was given in 2010[38] and 2011[39] at Plovdiv University, Bulgaria by Todor Arnaudov. MIT presented a course in AGI in 2018, organized by Lex Fridman and featuring a number of guest lecturers. However, as yet, most AI researchers have devoted little attention to AGI, with some claiming that intelligence is too complex to be completely replicated in the near term. However, a small number of computer scientists are active in AGI research, and many of this group are contributing to a series of AGI conferences. The research is extremely diverse and often pioneering in nature. In the introduction to his book,模板:Sfn Goertzel says that estimates of the time needed before a truly flexible AGI is built vary from 10 years to over a century, but the consensus in the AGI research community seems to be that the timeline discussed by Ray Kurzweil in The Singularity is Near[40] (i.e. between 2015 and 2045) is plausible.模板:Sfn

The term "artificial general intelligence" was used as early as 1997, by Mark Gubrud in a discussion of the implications of fully automated military production and operations. The term was re-introduced and popularized by Shane Legg and Ben Goertzel around 2002. The research objective is much older, for example Doug Lenat's Cyc project (that began in 1984), and Allen Newell's Soar project are regarded as within the scope of AGI. AGI research activity in 2006 was described by Pei Wang and Ben Goertzel as "producing publications and preliminary results". The first summer school in AGI was organized in Xiamen, China in 2009 by the Xiamen university's Artificial Brain Laboratory and OpenCog. The first university course was given in 2010 and 2011 at Plovdiv University, Bulgaria by Todor Arnaudov. MIT presented a course in AGI in 2018, organized by Lex Fridman and featuring a number of guest lecturers. However, as yet, most AI researchers have devoted little attention to AGI, with some claiming that intelligence is too complex to be completely replicated in the near term. However, a small number of computer scientists are active in AGI research, and many of this group are contributing to a series of AGI conferences. The research is extremely diverse and often pioneering in nature. In the introduction to his book, Goertzel says that estimates of the time needed before a truly flexible AGI is built vary from 10 years to over a century, but the consensus in the AGI research community seems to be that the timeline discussed by Ray Kurzweil in The Singularity is Near (i.e. between 2015 and 2045) is plausible.

“通用人工智能”一词早在1997年就由马克·古布鲁德(Mark Gubrud)在讨论全自动化军事生产和作业的影响时使用,又在2002年左右被肖恩·莱格(Shane Legg)和本·格兹尔(Ben Goertzel)重新引入并推广。通用人工智能的研究目标要古老得多,例如道格•雷纳特(Doug Lenat)的 Cyc 项目(始于1984年) ,以及艾伦•纽厄尔(Allen Newell)的 Soar 项目都被认为属于通用人工智能的范畴。王培(Pei Wang)和本·格兹尔将2006年的通用人工智能研究活动描述为“发表论文和取得初步成果”。2009年,厦门大学人工脑实验室和 OpenCog 在中国厦门组织了通用人工智能的第一个暑期学校。第一个大学课程于2010年和2011年在保加利亚普罗夫迪夫大学由托多尔·阿瑙多夫(Todor Arnaudov)开设。2018年,麻省理工学院开设了一门通用人工智能课程,由莱克斯·弗里德曼(Lex Fridman)组织,并邀请了一些客座讲师。然而,迄今为止,大多数人工智能研究人员对通用人工智能关注甚少,一些人声称,智能过于复杂,在短期内无法完全复制。然而,少数计算机科学家积极参与通用人工智能的研究,其中许多人正在为通用人工智能的一系列会议做出贡献。这项研究极其多样化,而且往往具有开创性。格兹尔在他的书的序言中,说,制造一个真正灵活的通用人工智能所需的时间约为10年到超过一个世纪不等,但是通用人工智能研究团体的似乎一致认为雷·库兹韦尔(Ray Kurzweil)在《奇点临近》(The Singularity is Near)(即在2015年至2045年之间)中讨论的时间线是可信的。


However, most mainstream AI researchers doubt that progress will be this rapid.[citation needed] Organizations explicitly pursuing AGI include the Swiss AI lab IDSIA,[citation needed] Nnaisense,[41] Vicarious, Maluuba,[7] the OpenCog Foundation, Adaptive AI, LIDA, and Numenta and the associated Redwood Neuroscience Institute.[42] In addition, organizations such as the Machine Intelligence Research Institute[43] and OpenAI[44] have been founded to influence the development path of AGI. Finally, projects such as the Human Brain Project[45] have the goal of building a functioning simulation of the human brain. A 2017 survey of AGI categorized forty-five known "active R&D projects" that explicitly or implicitly (through published research) research AGI, with the largest three being DeepMind, the Human Brain Project, and OpenAI.[7]

However, most mainstream AI researchers doubt that progress will be this rapid. Organizations explicitly pursuing AGI include the Swiss AI lab IDSIA, Nnaisense, Vicarious. In addition, organizations such as the Machine Intelligence Research Institute and OpenAI have been founded to influence the development path of AGI. Finally, projects such as the Human Brain Project have the goal of building a functioning simulation of the human brain. A 2017 survey of AGI categorized forty-five known "active R&D projects" that explicitly or implicitly (through published research) research AGI, with the largest three being DeepMind, the Human Brain Project, and OpenAI.

然而,大多数主流的人工智能研究人员怀疑进展是否会如此之快。明确寻求通用人工智能的组织包括瑞士人工智能实验室IDSIA,Nnaisense,Vicarious。此外,机器智能研究所和 OpenAI 等机构也建立起来以影响通用人工智能的发展道路。最后,还有像人脑计划这样的项目,目标是建立一个人脑的功能模拟。2017年针对一项通用人工智能的调查(通过已发表的研究)对45个已知的明确的或暗中研究通用人工智能的“活跃研发项目”进行了分类 ,其中最大的三个是 DeepMind、人类大脑项目和 OpenAI。


In 2017, researchers Feng Liu, Yong Shi and Ying Liu conducted intelligence tests on publicly available and freely accessible weak AI such as Google AI or Apple's Siri and others. At the maximum, these AI reached a value of about 47, which corresponds approximately to a six-year-old child in first grade. An adult comes to about 100 on average. Similar tests had been carried out in 2014, with the IQ score reaching a maximum value of 27.[46][47]

In 2017, researchers Feng Liu, Yong Shi and Ying Liu conducted intelligence tests on publicly available and freely accessible weak AI such as Google AI or Apple's Siri and others. At the maximum, these AI reached a value of about 47, which corresponds approximately to a six-year-old child in first grade. An adult comes to about 100 on average. Similar tests had been carried out in 2014, with the IQ score reaching a maximum value of 27.

2017年,研究人员刘锋、石勇和刘颖对公开的和可自由访问的弱人工智能进行了智商测试,如谷歌人工智能或苹果的 Siri 等。这些人工智能达到的最大值为约47,这大约相当于一个的六岁儿童。一个成年人平均智商为100。2014年也进行了类似的测试,智商分数的最高值达到了27。


In 2019, video game programmer and aerospace engineer John Carmack announced plans to research AGI.[48]

In 2019, video game programmer and aerospace engineer John Carmack announced plans to research AGI.

2019年,游戏程序师和航空工程师约翰·卡迈克(John Carmack)宣布了研究通用人工智能的计划。

模拟人脑所需要的处理能力

全脑模拟

A popular discussed approach to achieving general intelligent action is whole brain emulation. A low-level brain model is built by scanning and mapping a biological brain in detail and copying its state into a computer system or another computational device. The computer runs a simulation model so faithful to the original that it will behave in essentially the same way as the original brain, or for all practical purposes, indistinguishably.

A popular discussed approach to achieving general intelligent action is whole brain emulation. A low-level brain model is built by scanning and mapping a biological brain in detail and copying its state into a computer system or another computational device. The computer runs a simulation model so faithful to the original that it will behave in essentially the same way as the original brain, or for all practical purposes, indistinguishably.[49] Whole brain emulation is discussed in computational neuroscience and neuroinformatics, in the context of brain simulation for medical research purposes. It is discussed in artificial intelligence research模板:Sfn as an approach to strong AI. Neuroimaging technologies that could deliver the necessary detailed understanding are improving rapidly, and futurist Ray Kurzweil in the book The Singularity Is Near[40] predicts that a map of sufficient quality will become available on a similar timescale to the required computing power.

."The basic idea is to take a particular brain, scan its structure in detail, and construct a software model of it that is so faithful to the original that, when run on appropriate hardware, it will behave in essentially the same way as the original brain."</ref> Whole brain emulation is discussed in computational neuroscience and neuroinformatics, in the context of brain simulation for medical research purposes. It is discussed in artificial intelligence research as an approach to strong AI. Neuroimaging technologies that could deliver the necessary detailed understanding are improving rapidly, and futurist Ray Kurzweil in the book The Singularity Is Near predicts that a map of sufficient quality will become available on a similar timescale to the required computing power.

“基本思路是,取一个特定的大脑,详细地扫描其结构,并构建一个无比还原原始大脑的软件模型,以至于在适当的硬件上运行时,它基本上与原始大脑的行为方式相同。”在以医学研究为目的的背景下,全脑模拟在计算神经科学和神经信息学医学期刊上被讨论过。它是人工智能研究中讨论的一种强人工智能的方法。能够提供必要详细信息的神经成像技术正在迅速提高,而未来学家雷·库兹韦尔(Ray Kurzweil)在《奇点临近》书中预测,一张高质量的大脑图像将会出现,而同时,实现强人工智能所需的算力也会就绪。


早期预测

文件:Estimations of Human Brain Emulation Required Performance.svg
Estimates of how much processing power is needed to emulate a human brain at various levels (from Ray Kurzweil, and Anders Sandberg and Nick Bostrom), along with the fastest supercomputer from TOP500 mapped by year. Note the logarithmic scale and exponential trendline, which assumes the computational capacity doubles every 1.1 years. Kurzweil believes that mind uploading will be possible at neural simulation, while the Sandberg, Bostrom report is less certain about where consciousness arises.根据对在不同水平上模拟人类大脑的所需处理能力的估计(来自 Ray Kurzweil,[ Anders Sandberg 和 Nick Bostrom ]) ,以及每年从最快的五百台超级计算机获得的数据,绘制出对数尺度趋势线和指数趋势线。它呈现出计算能力每1.1年增长一倍。库兹韦尔相信,在神经模拟中上传思维是可能的,而桑德伯格和博斯特罗姆的报告对意识从何产生则不太确定。模板:Sfn

For low-level brain simulation, an extremely powerful computer would be required. The human brain has a huge number of synapses. Each of the 1011 (one hundred billion) neurons has on average 7,000 synaptic connections (synapses) to other neurons. It has been estimated that the brain of a three-year-old child has about 1015 synapses (1 quadrillion). This number declines with age, stabilizing by adulthood. Estimates vary for an adult, ranging from 1014 to 5×1014 synapses (100 to 500 trillion).模板:Sfn An estimate of the brain's processing power, based on a simple switch model for neuron activity, is around 1014 (100 trillion) synaptic updates per second (SUPS).模板:Sfn In 1997, Kurzweil looked at various estimates for the hardware required to equal the human brain and adopted a figure of 1016 computations per second (cps).[50] (For comparison, if a "computation" was equivalent to one "floating point operation" – a measure used to rate current supercomputers – then 1016 "computations" would be equivalent to 10 petaFLOPS, achieved in 2011). He used this figure to predict the necessary hardware would be available sometime between 2015 and 2025, if the exponential growth in computer power at the time of writing continued.

Estimates of how much processing power is needed to emulate a human brain at various levels (from Ray Kurzweil, and [[Anders Sandberg and Nick Bostrom), along with the fastest supercomputer from TOP500 mapped by year. Note the logarithmic scale and exponential trendline, which assumes the computational capacity doubles every 1.1 years. Kurzweil believes that mind uploading will be possible at neural simulation, while the Sandberg, Bostrom report is less certain about where consciousness arises.]] For low-level brain simulation, an extremely powerful computer would be required. The human brain has a huge number of synapses. Each of the 1011 (one hundred billion) neurons has on average 7,000 synaptic connections (synapses) to other neurons. It has been estimated that the brain of a three-year-old child has about 1015 synapses (1 quadrillion). This number declines with age, stabilizing by adulthood. Estimates vary for an adult, ranging from 1014 to 5×1014 synapses (100 to 500 trillion). An estimate of the brain's processing power, based on a simple switch model for neuron activity, is around 1014 (100 trillion) synaptic updates per second (SUPS). In 1997, Kurzweil looked at various estimates for the hardware required to equal the human brain and adopted a figure of 1016 computations per second (cps). (For comparison, if a "computation" was equivalent to one "floating point operation" – a measure used to rate current supercomputers – then 1016 "computations" would be equivalent to 10 petaFLOPS, achieved in 2011). He used this figure to predict the necessary hardware would be available sometime between 2015 and 2025, if the exponential growth in computer power at the time of writing continued.

文件:Estimations of Human Brain Emulation Required Performance.svg
根据对在不同水平上模拟人类大脑的所需处理能力的估计(来自 Ray Kurzweil,[ Anders Sandberg 和 Nick Bostrom ]) ,以及每年从最快的五百台超级计算机获得的数据,绘制出对数尺度趋势线和指数趋势线。它呈现出计算能力每1.1年增长一倍。库兹韦尔相信,在神经模拟中上传思维是可能的,而桑德伯格和博斯特罗姆的报告对意识从何产生则不太确定。

为进行低层次的大脑模拟,需要一个非常强大的计算机。人类的大脑有大量的突触。1011 (1000亿)个神经元中每一个平均与其他神经元有7000个突触连接(突触)。据估计,一个三岁儿童的大脑约有10 15 个突触(1千万亿)。这个数字随着年龄的增长而下降,成年后趋于稳定。而每个成年人的估计情况互不相同,从10个 14 到5 * 1014 个突触(100万亿到500万亿)不等。基于神经元活动的简单开关模型,对大脑处理能力的估计大约是每秒1014(100万亿)突触更新(SUPS)。1997年,库兹韦尔研究了等价模拟人脑所需硬件的各种估计,并采纳了每秒10 16 次 计算(cps)这个估计结果。(作为比较,如果一次“计算”相当于一次“浮点运算”——一种用于对当前超级计算机进行评级的措施——那么10 16 次“计算”相当于2011年达到的的每秒10000亿次浮点运算)。他用这个数字来预测,如果在撰写本文时计算机能力方面的指数增长继续下去的话,那么在2015年到2025年之间的某个时候,必要的硬件将会出现。


对神经元的更精细的模拟

The artificial neuron model assumed by Kurzweil and used in many current artificial neural network implementations is simple compared with biological neurons. A brain simulation would likely have to capture the detailed cellular behaviour of biological neurons, presently understood only in the broadest of outlines. The overhead introduced by full modeling of the biological, chemical, and physical details of neural behaviour (especially on a molecular scale) would require computational powers several orders of magnitude larger than Kurzweil's estimate. In addition the estimates do not account for glial cells, which are at least as numerous as neurons, and which may outnumber neurons by as much as 10:1, and are now known to play a role in cognitive processes.[51]

The artificial neuron model assumed by Kurzweil and used in many current artificial neural network implementations is simple compared with biological neurons. A brain simulation would likely have to capture the detailed cellular behaviour of biological neurons, presently understood only in the broadest of outlines. The overhead introduced by full modeling of the biological, chemical, and physical details of neural behaviour (especially on a molecular scale) would require computational powers several orders of magnitude larger than Kurzweil's estimate. In addition the estimates do not account for glial cells, which are at least as numerous as neurons, and which may outnumber neurons by as much as 10:1, and are now known to play a role in cognitive processes.

与生物神经元相比,库兹韦尔假设的人工神经元模型在当前许多人工神经网络(artificial neural network)实现中的应用是简单的。大脑模拟可能需要捕捉生物神经元细胞行为的细节,目前只能从最广泛的概要中理解。对神经行为的生物、化学和物理细节(特别是在分子尺度上)进行全面建模所需要的计算能力将比库兹韦尔的估计大数个数量级。此外,这些估计没有考虑到至少和神经元一样多的胶质细胞(glial cells),其数量可能是神经元的10倍,且现已知它们在认知过程中发挥作用。


研究现状

There are some research projects that are investigating brain simulation using more sophisticated neural models, implemented on conventional computing architectures. The Artificial Intelligence System project implemented non-real time simulations of a "brain" (with 1011 neurons) in 2005. It took 50 days on a cluster of 27 processors to simulate 1 second of a model.[52] The Blue Brain project used one of the fastest supercomputer architectures in the world, IBM's Blue Gene platform, to create a real time simulation of a single rat neocortical column consisting of approximately 10,000 neurons and 108 synapses in 2006.[53] A longer term goal is to build a detailed, functional simulation of the physiological processes in the human brain: "It is not impossible to build a human brain and we can do it in 10 years," Henry Markram, director of the Blue Brain Project said in 2009 at the TED conference in Oxford.[54] There have also been controversial claims to have simulated a cat brain. Neuro-silicon interfaces have been proposed as an alternative implementation strategy that may scale better.[55]

There are some research projects that are investigating brain simulation using more sophisticated neural models, implemented on conventional computing architectures. The Artificial Intelligence System project implemented non-real time simulations of a "brain" (with 1011 neurons) in 2005. It took 50 days on a cluster of 27 processors to simulate 1 second of a model. The Blue Brain project used one of the fastest supercomputer architectures in the world, IBM's Blue Gene platform, to create a real time simulation of a single rat neocortical column consisting of approximately 10,000 neurons and 108 synapses in 2006. A longer term goal is to build a detailed, functional simulation of the physiological processes in the human brain: "It is not impossible to build a human brain and we can do it in 10 years," Henry Markram, director of the Blue Brain Project said in 2009 at the TED conference in Oxford. There have also been controversial claims to have simulated a cat brain. Neuro-silicon interfaces have been proposed as an alternative implementation strategy that may scale better.

有一些研究项目正在使用更复杂的神经模型研究大脑模拟,这些模型是在传统的计算机体系结构上实现的。人工智能系统项目在2005年实现了对一个“大脑”(有10个 11 神经元)的非实时模拟。在一个由27个处理器组成的集群上,模拟一个模型的一秒钟花费了50天时间。2006年,蓝脑项目利用世界上最快的超级计算机架构之一——IBM 的蓝色基因平台,创建了一个包含大约10,000个神经元和10个 8 突触的单个大鼠的新皮质柱(neocortical column)的实时模拟。一个更长期的目标是建立一个人脑生理过程的详细的功能模拟: “建立一个人脑并不是不可能的,我们可以在10年内完成,”蓝脑项目主任亨利·马克拉姆(Henry Markram)于2009年在牛津举行的 TED 大会上说道。还有一些有争议的说法是模拟猫的大脑。神经-硅接口已作为一种可替代的实施策略被提出,它可能会更好地进行模拟。


Hans Moravec addressed the above arguments ("brains are more complicated", "neurons have to be modeled in more detail") in his 1997 paper "When will computer hardware match the human brain?".[56] He measured the ability of existing software to simulate the functionality of neural tissue, specifically the retina. His results do not depend on the number of glial cells, nor on what kinds of processing neurons perform where.

Hans Moravec addressed the above arguments ("brains are more complicated", "neurons have to be modeled in more detail") in his 1997 paper "When will computer hardware match the human brain?". He measured the ability of existing software to simulate the functionality of neural tissue, specifically the retina. His results do not depend on the number of glial cells, nor on what kinds of processing neurons perform where.

汉斯·莫拉维克(Hans Moravec)在他1997年的论文《计算机硬件何时能与人脑匹敌》中提出了上述观点(“大脑更复杂” ,“神经元的建模必须更详细”) .他测量了现有软件模拟神经组织,特别是视网膜功能的能力。他的研究结果并不取决于神经胶质细胞的数量,也不取决于处理神经元在哪里工作。


The actual complexity of modeling biological neurons has been explored in OpenWorm project that was aimed on complete simulation of a worm that has only 302 neurons in its neural network (among about 1000 cells in total). The animal's neural network has been well documented before the start of the project. However, although the task seemed simple at the beginning, the models based on a generic neural network did not work. Currently, the efforts are focused on precise emulation of biological neurons (partly on the molecular level), but the result cannot be called a total success yet. Even if the number of issues to be solved in a human-brain-scale model is not proportional to the number of neurons, the amount of work along this path is obvious.

The actual complexity of modeling biological neurons has been explored in OpenWorm project that was aimed on complete simulation of a worm that has only 302 neurons in its neural network (among about 1000 cells in total). The animal's neural network has been well documented before the start of the project. However, although the task seemed simple at the beginning, the models based on a generic neural network did not work. Currently, the efforts are focused on precise emulation of biological neurons (partly on the molecular level), but the result cannot be called a total success yet. Even if the number of issues to be solved in a human-brain-scale model is not proportional to the number of neurons, the amount of work along this path is obvious.

OpenWorm 项目已经探讨了建模生物神经元的实际复杂性。该项目旨在完全模拟一个蠕虫,其神经网络中只有302个神经元(在总共约1000个细胞中)。项目开始之前,蠕虫的神经网络已经被很好地记录了下来。然而,尽管任务一开始看起来很简单,基于一般神经网络的模型并不起作用。目前,研究的重点是精确模拟生物神经元(部分在分子水平上) ,但结果还不能被称为完全成功。即使在人脑尺度的模型中需要解决的问题的数量,与神经元的数量不成比例,沿着这条路径走下去的工作量也是显而易见的。


对基于模拟的研究方法的批评

A fundamental criticism of the simulated brain approach derives from embodied cognition where human embodiment is taken as an essential aspect of human intelligence. Many researchers believe that embodiment is necessary to ground meaning.[57] If this view is correct, any fully functional brain model will need to encompass more than just the neurons (i.e., a robotic body). Goertzel模板:Sfn proposes virtual embodiment (like in Second Life), but it is not yet known whether this would be sufficient.

A fundamental criticism of the simulated brain approach derives from embodied cognition where human embodiment is taken as an essential aspect of human intelligence. Many researchers believe that embodiment is necessary to ground meaning. If this view is correct, any fully functional brain model will need to encompass more than just the neurons (i.e., a robotic body). Goertzel proposes virtual embodiment (like in Second Life), but it is not yet known whether this would be sufficient.

对模拟大脑的方法的一个基本批评来自具象认知,其中人形被视为人类智力的一个重要方面。许多研究者认为,具象化是必要的基础意义。如果这种观点是正确的,那么任何功能齐全的大脑模型除了神经元还要包含更多东西(例如,一个机器人身体)。格兹尔提出了虚拟体(就像在《第二人生》中那样) ,但是目前还不知道这是否足够。


Desktop computers using microprocessors capable of more than 109 cps (Kurzweil's non-standard unit "computations per second", see above) have been available since 2005. According to the brain power estimates used by Kurzweil (and Moravec), this computer should be capable of supporting a simulation of a bee brain, but despite some interest[58] no such simulation exists [citation needed]. There are at least three reasons for this:

Desktop computers using microprocessors capable of more than 109 cps (Kurzweil's non-standard unit "computations per second", see above) have been available since 2005. According to the brain power estimates used by Kurzweil (and Moravec), this computer should be capable of supporting a simulation of a bee brain, but despite some interest no such simulation exists . There are at least three reasons for this:

自2005年以来,台式计算机使用的微处理器能够超过10 9 cps (库兹韦尔的非标准单位“每秒计算”,见上文)。根据库兹韦尔(和莫拉维克)使用的大脑能量估算,这台计算机应该能够支持蜜蜂大脑的模拟,但是尽管有些人感兴趣,这样的模拟却并不存在。这至少有三个原因:

  1. The neuron model seems to be oversimplified (see next section).

The neuron model seems to be oversimplified (see next section).


  1. There is insufficient understanding of higher cognitive processes模板:Refn to establish accurately what the brain's neural activity, observed using techniques such as functional magnetic resonance imaging, correlates with.

There is insufficient understanding of higher cognitive processes to establish accurately what the brain's neural activity, observed using techniques such as functional magnetic resonance imaging, correlates with.


  1. Even if our understanding of cognition advances sufficiently, early simulation programs are likely to be very inefficient and will, therefore, need considerably more hardware.

Even if our understanding of cognition advances sufficiently, early simulation programs are likely to be very inefficient and will, therefore, need considerably more hardware.


  1. The brain of an organism, while critical, may not be an appropriate boundary for a cognitive model. To simulate a bee brain, it may be necessary to simulate the body, and the environment. The Extended Mind thesis formalizes the philosophical concept, and research into cephalopods has demonstrated clear examples of a decentralized system.[59]

The brain of an organism, while critical, may not be an appropriate boundary for a cognitive model. To simulate a bee brain, it may be necessary to simulate the body, and the environment. The Extended Mind thesis formalizes the philosophical concept, and research into cephalopods has demonstrated clear examples of a decentralized system.


  1. 神经元模型似乎被过于简化了(见下一节)。
  2. 即使我们对认知的理解有了足够的进步,早期的仿真程序也可能非常低效,因此需要更多的硬件。
  3. 人们对高级认知过程的理解不够充分,使用功能性磁共振成像等技术观察到的大脑活动无法让人们准确地确定大脑的神经活动。
  4. 有机体的大脑虽然关键,但可能不是认知模型的合适边界。为了模拟蜜蜂的大脑,可能需要模拟身体和环境。延展心灵论题(The Extended Mind thesis)形式化了哲学概念,对头足类动物的研究已经展示了分散系统的明显的例子。


In addition, the scale of the human brain is not currently well-constrained. One estimate puts the human brain at about 100 billion neurons and 100 trillion synapses.[60][61] Another estimate is 86 billion neurons of which 16.3 billion are in the cerebral cortex and 69 billion in the cerebellum.模板:Sfn Glial cell synapses are currently unquantified but are known to be extremely numerous.

In addition, the scale of the human brain is not currently well-constrained. One estimate puts the human brain at about 100 billion neurons and 100 trillion synapses. Another estimate is 86 billion neurons of which 16.3 billion are in the cerebral cortex and 69 billion in the cerebellum. Glial cell synapses are currently unquantified but are known to be extremely numerous.

此外,人类大脑的规模上限目前还没有得到很好的估计。据测算,人类大脑大约有1000亿个神经元和100万亿个突触。另一个估计是860亿个神经元,其中163亿个在大脑皮层,690亿个在小脑。神经胶质细胞突触目前尚未定量,但已知数量极多。

强人工智能和意识


In 1980, philosopher John Searle coined the term "strong AI" as part of his Chinese room argument.[62] He wanted to distinguish between two different hypotheses about artificial intelligence:[63]

In 1980, philosopher John Searle coined the term "strong AI" as part of his Chinese room argument. He wanted to distinguish between two different hypotheses about artificial intelligence:

1980年,哲学家约翰•塞尔(John Searle)提出“强人工智能”(strong AI)一词作为他在中文屋论证的一部分。他想要区分关于人工智能的两种不同假设:

  • An artificial intelligence system can think and have a mind. (The word "mind" has a specific meaning for philosophers, as used in "the mind body problem" or "the philosophy of mind".)
  • 一个人工智能系统可以思考并拥有心灵。(词语“心灵”对哲学家来说有特殊意义,正如在“身心问题”或“心灵哲学”中的使用一样。)
  • An artificial intelligence system can (only) act like it thinks and has a mind.
  • 一个人工智能系统(仅仅)可以表现得像是能思考、拥有心灵那样去行动。

The first one is called "the strong AI hypothesis" and the second is "the weak AI hypothesis" because the first one makes the stronger statement: it assumes something special has happened to the machine that goes beyond all its abilities that we can test. Searle referred to the "strong AI hypothesis" as "strong AI". This usage is also common in academic AI research and textbooks.

The first one is called "the strong AI hypothesis" and the second is "the weak AI hypothesis" because the first one makes the stronger statement: it assumes something special has happened to the machine that goes beyond all its abilities that we can test. Searle referred to the "strong AI hypothesis" as "strong AI". This usage is also common in academic AI research and textbooks.

第一条被称为“强人工智能假设(the strong AI hypothesis)” ,第二条被称为“弱人工智能假设(the weak AI hypothesis)”,因为第一条假设提出了更强的陈述: 它假定机器发生了某种特殊的事件,超出了我们能够测试的所有能力。塞尔将“强人工智能假说”称为“强人工智能”。这种用法在人工智能学术研究和教科书中也很常见。[64]


The weak AI hypothesis is equivalent to the hypothesis that artificial general intelligence is possible. According to Russell and Norvig, "Most AI researchers take the weak AI hypothesis for granted, and don't care about the strong AI hypothesis."模板:Sfn

The weak AI hypothesis is equivalent to the hypothesis that artificial general intelligence is possible. According to Russell and Norvig, "Most AI researchers take the weak AI hypothesis for granted, and don't care about the strong AI hypothesis."

弱人工智能假说其实就意味着通用人工智能是可能的。根据罗素和诺维格的说法,“大多数人工智能研究人员认为弱人工智能假说是理所当然的,并且不关心强人工智能假说。”


In contrast to Searle, Ray Kurzweil uses the term "strong AI" to describe any artificial intelligence system that acts like it has a mind,[40] regardless of whether a philosopher would be able to determine if it actually has a mind or not.

In contrast to Searle, Ray Kurzweil uses the term "strong AI" to describe any artificial intelligence system that acts like it has a mind, regardless of whether a philosopher would be able to determine if it actually has a mind or not.

与 Searle 不同的是,Ray Kurzweil 使用“强人工智能”这个词来描述任何人工智能系统,这个系统的行为就像它有思想一样,不管哲学家能否确定它是否真的有思想。

In science fiction, AGI is associated with traits such as consciousness, sentience, sapience, and self-awareness observed in living beings. However, according to Searle, it is an open question whether general intelligence is sufficient for consciousness. "Strong AI" (as defined above by Kurzweil) should not be confused with Searle's "strong AI hypothesis." The strong AI hypothesis is the claim that a computer which behaves as intelligently as a person must also necessarily have a mind and consciousness. AGI refers only to the amount of intelligence that the machine displays, with or without a mind.

In science fiction, AGI is associated with traits such as consciousness, sentience, sapience, and self-awareness observed in living beings. However, according to Searle, it is an open question whether general intelligence is sufficient for consciousness. "Strong AI" (as defined above by Kurzweil) should not be confused with Searle's "strong AI hypothesis." The strong AI hypothesis is the claim that a computer which behaves as intelligently as a person must also necessarily have a mind and consciousness. AGI refers only to the amount of intelligence that the machine displays, with or without a mind.

在科幻小说中,通用人工智能与生物所具有的的意识、知觉、智慧和自我意识等特征有关。然而,根据塞尔的说法,一般智力是否足以产生意识还是一个悬而未决的问题。“强人工智能”(如上文库兹韦尔所定义的)不应与塞尔的“强人工智能假设”相混淆。强人工智能假说认为,一台像人一样智能运行的计算机必然具有思想和意识。通用人工智能只是指机器显示出来的智能程度,而与机器是否拥有心灵无关。


意识

There are other aspects of the human mind besides intelligence that are relevant to the concept of strong AI which play a major role in science fiction and the ethics of artificial intelligence:

There are other aspects of the human mind besides intelligence that are relevant to the concept of strong AI which play a major role in science fiction and the ethics of artificial intelligence:

除了“智能”这一与强人工智能概念相关的方面,人类思维还有其他方面:


  • 意识:拥有主观体验和思想。值得一提的是意识是很难定义的。由托马斯·内格尔给出的一个著名定义陈述如下:一个事物如果能体会到某种感觉,那么它是有意识的。如果我们不是有意识的,那么我们不会有任何感觉。内格尔以蝙蝠为例:我们可以凭借感觉问出:“成为一只蝙蝠的感觉如何?”但是,我们不大可能问出:“成为一个吐司机的感觉如何?”内格尔总结认为蝙蝠像是有意识的(即拥有意识),但是吐司机却不是。
  • 自我意识:能够意识到自己是一个独立的个体,尤其是意识到自己的思想。
  • 知觉:主观地感受概念或者情感的能力。
  • 智慧:容纳知识的能力。


These traits have a moral dimension, because a machine with this form of strong AI may have legal rights, analogous to the rights of non-human animals. As such, preliminary work has been conducted on approaches to integrating full ethical agents with existing legal and social frameworks. These approaches have focused on the legal position and rights of 'strong' AI.[66]

These traits have a moral dimension, because a machine with this form of strong AI may have legal rights, analogous to the rights of non-human animals. As such, preliminary work has been conducted on approaches to integrating full ethical agents with existing legal and social frameworks. These approaches have focused on the legal position and rights of 'strong' AI.

这些特征具有道德维度,因为拥有这种强人工智能形式的机器可能拥有法律权利,类似于非人类动物的权利。因此,一些初步工作已经开展,探讨如何将完全道德主体纳入现有的法律和社会框架。这些方法都集中在“强”人工智能的法律地位和权利上。


However, Bill Joy, among others, argues a machine with these traits may be a threat to human life or dignity.[67] It remains to be shown whether any of these traits are necessary for strong AI. The role of consciousness is not clear, and currently there is no agreed test for its presence. If a machine is built with a device that simulates the neural correlates of consciousness, would it automatically have self-awareness? It is also possible that some of these properties, such as sentience, naturally emerge from a fully intelligent machine, or that it becomes natural to ascribe these properties to machines once they begin to act in a way that is clearly intelligent. For example, intelligent action may be sufficient for sentience, rather than the other way around.

However, Bill Joy, among others, argues a machine with these traits may be a threat to human life or dignity. It remains to be shown whether any of these traits are necessary for strong AI. The role of consciousness is not clear, and currently there is no agreed test for its presence. If a machine is built with a device that simulates the neural correlates of consciousness, would it automatically have self-awareness? It is also possible that some of these properties, such as sentience, naturally emerge from a fully intelligent machine, or that it becomes natural to ascribe these properties to machines once they begin to act in a way that is clearly intelligent. For example, intelligent action may be sufficient for sentience, rather than the other way around.

然而,比尔·乔伊(Bill Joy)等人认为,具有这些特征的机器可能会威胁到人类的生命或尊严。这些特征对于强人工智能来说是否是必要的还有待证明。意识的作用并不清楚,目前也没有公认的测试来确定其存在。如果一台机器装有一个能模拟与意识相关神经的装置,它会自动具有自我意识吗?也有可能这些特性中的一些,比如感知能力,自然而然地从一个完全智能的机器中产生,或者一旦机器开始以一种明显智能的方式行动,人们就会自然而然地认为这些特性是机器自主产生的。例如,智能行为可能足以判定机器产生了知觉,而非反过来。

 --袁一博讨论)“或者一旦机器开始以一种明显智能的方式行动,,人们就会自然而然地认为这些特性是机器自主产生的。”对应原句“or that it becomes natural to ascribe these properties to machines once they begin to act in a way that is clearly intelligent.”与原句在语序和措辞上略有不同,是译者考虑到中文的阅读习惯在不改变原意的条件下意译得出的。
 --Ricky讨论)已采纳

人工意识研究

Although the role of consciousness in strong AI/AGI is debatable, many AGI researchers模板:Sfn regard research that investigates possibilities for implementing consciousness as vital. In an early effort Igor Aleksander模板:Sfn argued that the principles for creating a conscious machine already existed but that it would take forty years to train such a machine to understand language.

Although the role of consciousness in strong AI/AGI is debatable, many AGI researchers regard research that investigates possibilities for implementing consciousness as vital. In an early effort Igor Aleksander argued that the principles for creating a conscious machine already existed but that it would take forty years to train such a machine to understand language.

虽然意识在强人工智能/通用人工智能中的作用是有争议的,但是很多通用人工智能的研究人员认为研究实现意识的可能性是至关重要的。在早期的努力中,伊格尔·亚历山大(Igor Aleksander)认为如何创造一个有意识的机器的知识已经存在,但是训练这样一个机器去理解语言可能需要四十年。

对于人工智能研究进展缓慢的可能解释


Since the launch of AI research in 1956, the growth of this field has slowed down over time and has stalled the aims of creating machines skilled with intelligent action at the human level.模板:Sfn A possible explanation for this delay is that computers lack a sufficient scope of memory or processing power.模板:Sfn In addition, the level of complexity that connects to the process of AI research may also limit the progress of AI research.模板:Sfn

Since the launch of AI research in 1956, the growth of this field has slowed down over time and has stalled the aims of creating machines skilled with intelligent action at the human level. A possible explanation for this delay is that computers lack a sufficient scope of memory or processing power. In addition, the level of complexity that connects to the process of AI research may also limit the progress of AI research.

自从1956年人工智能研究启动以来,这一领域的发展速度已经随着时间的推移而放缓,创造具有人类水平智能机器的目标依然遥不可及。这种延迟的一个可能的解释是计算机缺乏足够的存储空间或处理能力。此外,人工智能研究过程的复杂程度也可能限制人工智能研究的进展。


While most AI researchers believe strong AI can be achieved in the future, there are some individuals like Hubert Dreyfus and Roger Penrose who deny the possibility of achieving strong AI.模板:Sfn John McCarthy was one of various computer scientists who believe human-level AI will be accomplished, but a date cannot accurately be predicted.模板:Sfn

While most AI researchers believe strong AI can be achieved in the future, there are some individuals like Hubert Dreyfus and Roger Penrose who deny the possibility of achieving strong AI. John McCarthy was one of various computer scientists who believe human-level AI will be accomplished, but a date cannot accurately be predicted.

虽然大多数人工智能研究人员认为强人工智能可以在未来实现,但也有一些像休伯特·德雷福斯(Hubert Dreyfus)和罗杰·彭罗斯(Roger Penrose)那样否认实现强人工智能的可能性。约翰·麦卡锡(John McCarthy)是众多相信人类水平人工智能将会实现的计算机科学家之一,但是日期无法准确预测。


Conceptual limitations are another possible reason for the slowness in AI research.模板:Sfn AI researchers may need to modify the conceptual framework of their discipline in order to provide a stronger base and contribution to the quest of achieving strong AI. As William Clocksin wrote in 2003: "the framework starts from Weizenbaum's observation that intelligence manifests itself only relative to specific social and cultural contexts".模板:Sfn

Conceptual limitations are another possible reason for the slowness in AI research. AI researchers may need to modify the conceptual framework of their discipline in order to provide a stronger base and contribution to the quest of achieving strong AI. As William Clocksin wrote in 2003: "the framework starts from Weizenbaum's observation that intelligence manifests itself only relative to specific social and cultural contexts".

概念上的局限性是人工智能研究缓慢的另一个可能原因。人工智能研究人员可能需要修改他们学科的概念框架,以便为实现强人工智能提供一个更强大的基础。正如威廉·克罗克森(William Clocksin)在2003年写的那样: “这个框架始于魏岑鲍姆(Weizenbaum)的观察,即智能只在特定的社会文化背景下才能表现出来。”。


Furthermore, AI researchers have been able to create computers that can perform jobs that are complicated for people to do, such as mathematics, but conversely they have struggled to develop a computer that is capable of carrying out tasks that are simple for humans to do, such as walking (Moravec's paradox).模板:Sfn A problem described by David Gelernter is that some people assume thinking and reasoning are equivalent.模板:Sfn However, the idea of whether thoughts and the creator of those thoughts are isolated individually has intrigued AI researchers.模板:Sfn

Furthermore, AI researchers have been able to create computers that can perform jobs that are complicated for people to do, such as mathematics, but conversely they have struggled to develop a computer that is capable of carrying out tasks that are simple for humans to do, such as walking (Moravec's paradox). A problem described by David Gelernter is that some people assume thinking and reasoning are equivalent. However, the idea of whether thoughts and the creator of those thoughts are isolated individually has intrigued AI researchers.

此外,人工智能研究人员已经能够创造出能够执行对人类而言复杂的工作(如数学)的计算机,但相反地,他们却难以开发出能够执行人类简单任务(如行走)的计算机(莫拉维克悖论(Moravec's paradox))。大卫·格勒尼特(David Gelernter)描述的一个问题是,有些人认为思考和推理是等价的。然而,人工智能研究者已经开始探究是否思想和这些思想的创造者是分开的。


The problems that have been encountered in AI research over the past decades have further impeded the progress of AI. The failed predictions that have been promised by AI researchers and the lack of a complete understanding of human behaviors have helped diminish the primary idea of human-level AI.模板:Sfn Although the progress of AI research has brought both improvement and disappointment, most investigators have established optimism about potentially achieving the goal of AI in the 21st century.模板:Sfn

The problems that have been encountered in AI research over the past decades have further impeded the progress of AI. The failed predictions that have been promised by AI researchers and the lack of a complete understanding of human behaviors have helped diminish the primary idea of human-level AI. Although the progress of AI research has brought both improvement and disappointment, most investigators have established optimism about potentially achieving the goal of AI in the 21st century.

过去几十年人工智能研究中遇到的问题进一步阻碍了人工智能的发展。人工智能研究人员所做的失败预测,以及对人类行为不够完整的理解,削弱了人类水平人工智能的最初设想。尽管人工智能研究的进展带来了进步和失望,但大多数研究人员对人工智能在21世纪可能实现的目标持乐观态度。


Other possible reasons have been proposed for the lengthy research in the progress of strong AI. The intricacy of scientific problems and the need to fully understand the human brain through psychology and neurophysiology have limited many researchers in emulating the function of the human brain in computer hardware.模板:Sfn Many researchers tend to underestimate any doubt that is involved with future predictions of AI, but without taking those issues seriously, people can then overlook solutions to problematic questions.模板:Sfn

Other possible reasons have been proposed for the lengthy research in the progress of strong AI. The intricacy of scientific problems and the need to fully understand the human brain through psychology and neurophysiology have limited many researchers in emulating the function of the human brain in computer hardware. Many researchers tend to underestimate any doubt that is involved with future predictions of AI, but without taking those issues seriously, people can then overlook solutions to problematic questions.

还有其他可能的原因也能解释为什么强人工智能的研究持续了这么长时间。错综复杂的科学问题,以及用心理学和神经生理学充分了解人脑的需要,限制了许多研究人员在计算机硬件中模拟人脑。许多研究人员倾向于低估对人工智能未来预测有关的任何怀疑,但是如果不认真对待这些问题,人们就会忽视这些问题的解决方案。


Clocksin says that a conceptual limitation that may impede the progress of AI research is that people may be using the wrong techniques for computer programs and implementation of equipment.模板:Sfn When AI researchers first began to aim for the goal of artificial intelligence, a main interest was human reasoning.模板:Sfn Researchers hoped to establish computational models of human knowledge through reasoning and to find out how to design a computer with a specific cognitive task.模板:Sfn

Clocksin says that a conceptual limitation that may impede the progress of AI research is that people may be using the wrong techniques for computer programs and implementation of equipment. When AI researchers first began to aim for the goal of artificial intelligence, a main interest was human reasoning. Researchers hoped to establish computational models of human knowledge through reasoning and to find out how to design a computer with a specific cognitive task.

克罗克森说,阻碍人工智能研究进展的一个概念上的限制是,人们可能在计算机程序和安装设备方面使用了错误的技术。当人工智能研究人员第一次开始瞄准人工智能的目标时,主要的兴趣是人类推理。研究人员希望通过推理建立人类知识的计算模型,并找出如何设计一台具有特定认知任务的计算机。


The practice of abstraction, which people tend to redefine when working with a particular context in research, provides researchers with a concentration on just a few concepts.模板:Sfn The most productive use of abstraction in AI research comes from planning and problem solving.模板:Sfn Although the aim is to increase the speed of a computation, the role of abstraction has posed questions about the involvement of abstraction operators.模板:Sfn

The practice of abstraction, which people tend to redefine when working with a particular context in research, provides researchers with a concentration on just a few concepts. The most productive use of abstraction in AI research comes from planning and problem solving. Although the aim is to increase the speed of a computation, the role of abstraction has posed questions about the involvement of abstraction operators.

人们倾向于在特定研究领域中使用抽象,这使得研究人员能集中在少数几个概念上。抽象在人工智能研究中最有效的应用来自规划和解决问题。虽然目标是提高计算速度,但是抽象的作用和角色已经在呼唤一种能用哦股计算的抽象操作。


A possible reason for the slowness in AI relates to the acknowledgement by many AI researchers that heuristics is a section that contains a significant breach between computer performance and human performance.模板:Sfn The specific functions that are programmed to a computer may be able to account for many of the requirements that allow it to match human intelligence. These explanations are not necessarily guaranteed to be the fundamental causes for the delay in achieving strong AI, but they are widely agreed by numerous researchers.

A possible reason for the slowness in AI relates to the acknowledgement by many AI researchers that heuristics is a section that contains a significant breach between computer performance and human performance. The specific functions that are programmed to a computer may be able to account for many of the requirements that allow it to match human intelligence. These explanations are not necessarily guaranteed to be the fundamental causes for the delay in achieving strong AI, but they are widely agreed by numerous researchers.

 -- Ricky(讨论)这段英文审校也读不懂。

人工智能发展缓慢的一个可能原因是许多人工智能研究人员承认启发式算法是计算机性能和人类表现之间的一个重大缺口。为计算机编入的特定功能可以满足许多要求,使计算机与人类智能相匹配。这些解释并不一定是延迟人工智能的实现的根本原因,但它们得到了众多研究人员的广泛认同。


There have been many AI researchers that debate over the idea whether machines should be created with emotions. There are no emotions in typical models of AI and some researchers say programming emotions into machines allows them to have a mind of their own.模板:Sfn Emotion sums up the experiences of humans because it allows them to remember those experiences.模板:Sfn David Gelernter writes, "No computer will be creative unless it can simulate all the nuances of human emotion."模板:Sfn This concern about emotion has posed problems for AI researchers and it connects to the concept of strong AI as its research progresses into the future.[68]

There have been many AI researchers that debate over the idea whether machines should be created with emotions. There are no emotions in typical models of AI and some researchers say programming emotions into machines allows them to have a mind of their own. Emotion sums up the experiences of humans because it allows them to remember those experiences. David Gelernter writes, "No computer will be creative unless it can simulate all the nuances of human emotion." This concern about emotion has posed problems for AI researchers and it connects to the concept of strong AI as its research progresses into the future.

许多人工智能研究人员一直在争论机器是否应该带有情感。典型的人工智能模型中没有情感,一些研究人员说,将情感编程到机器中可以让它们拥有自己的思想。情感总结了人类的经历,它使得人们记住那些经历。大卫·格勒尼特(David Gelernter)则写道: “除非计算机能够模拟人类情感的所有细微差别,否则它不会具有创造力。”这种对情绪的关注给人工智能研究人员带来了一些问题,随着未来人工智能研究的进展,它与强人工智能的概念联系起来。

Controversies and dangers 争议和风险

Feasibility 可能性

模板:Expand section

As of March 2020, AGI remains speculative[69][70] as no such system has been demonstrated yet. Opinions vary both on whether and when artificial general intelligence will arrive. At one extreme, AI pioneer Herbert A. Simon wrote in 1965: "machines will be capable, within twenty years, of doing any work a man can do". However, this prediction failed to come true. Microsoft co-founder Paul Allen believed that such intelligence is unlikely in the 21st century because it would require "unforeseeable and fundamentally unpredictable breakthroughs" and a "scientifically deep understanding of cognition".[71] Writing in The Guardian, roboticist Alan Winfield claimed the gulf between modern computing and human-level artificial intelligence is as wide as the gulf between current space flight and practical faster-than-light spaceflight.[72]

As of March 2020, AGI remains speculative as no such system has been demonstrated yet. Opinions vary both on whether and when artificial general intelligence will arrive. At one extreme, AI pioneer Herbert A. Simon wrote in 1965: "machines will be capable, within twenty years, of doing any work a man can do". However, this prediction failed to come true. Microsoft co-founder Paul Allen believed that such intelligence is unlikely in the 21st century because it would require "unforeseeable and fundamentally unpredictable breakthroughs" and a "scientifically deep understanding of cognition". Writing in The Guardian, roboticist Alan Winfield claimed the gulf between modern computing and human-level artificial intelligence is as wide as the gulf between current space flight and practical faster-than-light spaceflight.

截至2020年3月,通用人工智能仍处于推测性的状态,因为迄今此类系统尚未被展示。对于人工通用智能是否会到来以及何时到来,人们的看法各不相同。一个极端如,人工智能的先驱赫伯特·西蒙在1965年写道: “机器将能在20年内具有完成人类能做的任何工作的能力。”然而,这个预言并没有实现。微软(Microsoft)联合创始人保罗•艾伦(Paul Allen)认为,这种情报在21世纪不太可能出现,因为它需要“不可预见且根本无法预测的突破”和“对认知的科学的深入理解”。机器人专家阿兰·温菲尔德(Alan Winfield)在《卫报》上发表文章称,现代计算机和人类水平的人工智能之间的鸿沟就像当前的太空飞行和实际的超光速空间飞行之间的鸿沟一样宽。


AI experts' views on the feasibility of AGI wax and wane, and may have seen a resurgence in the 2010s. Four polls conducted in 2012 and 2013 suggested that the median guess among experts for when they would be 50% confident AGI would arrive was 2040 to 2050, depending on the poll, with the mean being 2081. Of the experts, 16.5% answered with "never" when asked the same question but with a 90% confidence instead.[73][74] Further current AGI progress considerations can be found below Tests for confirming human-level AGI and IQ-tests AGI.

AI experts' views on the feasibility of AGI wax and wane, and may have seen a resurgence in the 2010s. Four polls conducted in 2012 and 2013 suggested that the median guess among experts for when they would be 50% confident AGI would arrive was 2040 to 2050, depending on the poll, with the mean being 2081. Of the experts, 16.5% answered with "never" when asked the same question but with a 90% confidence instead. Further current AGI progress considerations can be found below Tests for confirming human-level AGI and IQ-tests AGI.

人工智能专家一直在考虑通用人工智能兴衰可能性,可能在2010年代出现了复苏。2012年和2013年进行的四次民意调查显示,专家们平均有50%的信心相信通用人工智能会在2040年到2050年间到来,具体取决于调查结果,平均猜测是2081年。在这些专家中,16.5% 的人在被问到同样的问题时回答“从来没有” ,但他们的自信心却达到了90%。当前通用人工智能项目进展中的进一步考虑可以在确认人类水平通用人工智能和基于智商测试的通用人工智能测试下面找到。


对人类的潜在威胁

The thesis that AI poses an existential risk, and that this risk needs much more attention than it currently gets, has been endorsed by many public figures; perhaps the most famous are Elon Musk, Bill Gates, and Stephen Hawking. The most notable AI researcher to endorse the thesis is Stuart J. Russell. Endorsers of the thesis sometimes express bafflement at skeptics: Gates states he does not "understand why some people are not concerned",[75] and Hawking criticized widespread indifference in his 2014 editorial: 模板:Cquote

The thesis that AI poses an existential risk, and that this risk needs much more attention than it currently gets, has been endorsed by many public figures; perhaps the most famous are Elon Musk, Bill Gates, and Stephen Hawking. The most notable AI researcher to endorse the thesis is Stuart J. Russell. Endorsers of the thesis sometimes express bafflement at skeptics: Gates states he does not "understand why some people are not concerned", and Hawking criticized widespread indifference in his 2014 editorial: we'll leave the lights on?' Probably not, but this is more or less what is happening with AI.'}}

一篇论文提出了“人工智能可能会引发世界末日”,且这种风险需要更多关注。这一论点也已经得到了许多公众人物的支持; 也许最著名的是埃隆·马斯克,比尔·盖茨和斯蒂芬·霍金。支持这一观点的最著名的人工智能研究者是斯图尔特·罗素。这篇论文的支持者有时会对怀疑论者表示困惑: 盖茨表示,他不“理解为什么有些人不关心” ,霍金在2014年的社论中批评了普遍的冷漠: 我们就这样对这个问题不管不顾吗?可能不是,但这或多或少是正在发生在人工智能领域内的。


Many of the scholars who are concerned about existential risk believe that the best way forward would be to conduct (possibly massive) research into solving the difficult "control problem" to answer the question: what types of safeguards, algorithms, or architectures can programmers implement to maximize the probability that their recursively-improving AI would continue to behave in a friendly, rather than destructive, manner after it reaches superintelligence?[76][77]

Many of the scholars who are concerned about existential risk believe that the best way forward would be to conduct (possibly massive) research into solving the difficult "control problem" to answer the question: what types of safeguards, algorithms, or architectures can programmers implement to maximize the probability that their recursively-improving AI would continue to behave in a friendly, rather than destructive, manner after it reaches superintelligence?

许多关注末日风险的学者认为,最好的方法是进行(可能是大规模的)研究,去解决困难的“控制问题”,以回答这个疑问: 程序员可以实现哪些保障措施、算法或架构,以最大程度地确保其不断改进的人工智能在达到超级智能后会继续以友好地运行,而非破坏?


The thesis that AI can pose existential risk also has many strong detractors. Skeptics sometimes charge that the thesis is crypto-religious, with an irrational belief in the possibility of superintelligence replacing an irrational belief in an omnipotent God; at an extreme, Jaron Lanier argues that the whole concept that current machines are in any way intelligent is "an illusion" and a "stupendous con" by the wealthy.[78]

The thesis that AI can pose existential risk also has many strong detractors. Skeptics sometimes charge that the thesis is crypto-religious, with an irrational belief in the possibility of superintelligence replacing an irrational belief in an omnipotent God; at an extreme, Jaron Lanier argues that the whole concept that current machines are in any way intelligent is "an illusion" and a "stupendous con" by the wealthy.

人工智能可以造成世界末日的观点也遭到了许多强烈的反对。怀疑论者有时指责该论点是神秘的、宗教性的、非理性地相信超级智能可能取代对万能的上帝的非理性信仰; 极端如杰伦·拉尼尔(Jaron Lanier)认为,目前机器以任何方式具有智能只是“一种幻觉” ,是富人的“惊人骗局”。


Much of existing criticism argues that AGI is unlikely in the short term. Computer scientist Gordon Bell argues that the human race will already destroy itself before it reaches the technological singularity. Gordon Moore, the original proponent of Moore's Law, declares that "I am a skeptic. I don't believe [a technological singularity] is likely to happen, at least for a long time. And I don't know why I feel that way."[79] Baidu Vice President Andrew Ng states AI existential risk is "like worrying about overpopulation on Mars when we have not even set foot on the planet yet."[80]

Much of existing criticism argues that AGI is unlikely in the short term. Computer scientist Gordon Bell argues that the human race will already destroy itself before it reaches the technological singularity. Gordon Moore, the original proponent of Moore's Law, declares that "I am a skeptic. I don't believe [a technological singularity] is likely to happen, at least for a long time. And I don't know why I feel that way." Baidu Vice President Andrew Ng states AI existential risk is "like worrying about overpopulation on Mars when we have not even set foot on the planet yet."

现有的许多批评认为,通用人工智能短期内不太可能成功。计算机科学家戈登·贝尔(Gordon Bell)认为人类在到达技术奇点(technological singularity)之前就已经自我毁灭了。而戈登·摩尔(Gordon Moore)——摩尔定律(Moore's Law)的最初提出者,宣称“我是一个怀疑论者。我不认为技术奇点会发生,至少在很长一段时间内不会。我不知道为什么会有这种感觉。”百度副总裁吴恩达(Andrew Ng)说,人工智能世界末日就像是在担心火星人口过剩,而我们甚至还没有踏上这个星球。

See also 请参阅

  • Intelligence amplification (IA), the use of information technology in augmenting human intelligence instead of creating an external autonomous "AGI"

智能放大,利用信息技术加强人类智慧而不是建造外在的通用人工智能


Notes 附注

  1. "DeepMind and Google: the battle to control artificial intelligence". The Economist (1843 (magazine)). 2019. Retrieved 15 March 2020. AGI stands for Artificial General Intelligence, a hypothetical computer program...
  2. Kurzweil, Singularity (2005) p. 260
  3. 3.0 3.1 Kurzweil, Ray (5 August 2005), "Long Live AI", Forbes: Kurzweil describes strong AI as "machine intelligence with the full range of human intelligence."
  4. "DeepMind and Google: the battle to control artificial intelligence". The Economist (1843 (magazine)). 2019. Retrieved 15 March 2020. AGI stands for Artificial General Intelligence, a hypothetical computer program...
  5. Encyclopædia Britannica Strong AI, applied AI, and cognitive simulation -{zh-cn:互联网档案馆; zh-tw:網際網路檔案館; zh-hk:互聯網檔案館;}-存檔,存档日期15 October 2007. or Jack Copeland What is artificial intelligence? -{zh-cn:互联网档案馆; zh-tw:網際網路檔案館; zh-hk:互聯網檔案館;}-存檔,存档日期18 August 2007. on AlanTuring.net
  6. "The Open University on Strong and Weak AI". Archived from the original on 25 September 2009. Retrieved 8 October 2007. https://en.wikipedia.org/wiki/Defekte_Weblinks?dwl={{{url}}} Seite nicht mehr abrufbar], Suche in Webarchiven: Kategorie:Wikipedia:Weblink offline (andere Namensräume)[http://timetravel.mementoweb.org/list/2010/Kategorie:Wikipedia:Vorlagenfehler/Vorlage:Toter Link/URL_fehlt
  7. 7.0 7.1 7.2 Baum, Seth (12 November 2017). "Baum, Seth, A Survey of Artificial General Intelligence Projects for Ethics, Risk, and Policy (November 12, 2017). Global Catastrophic Risk Institute Working Paper 17-1". Cite journal requires |journal= (help)
  8. AI founder John McCarthy writes: "we cannot yet characterize in general what kinds of computational procedures we want to call intelligent." McCarthy, John (2007). "Basic Questions". Stanford University. Archived from the original on 26 October 2007. Retrieved 6 December 2007. (For a discussion of some definitions of intelligence used by artificial intelligence researchers, see philosophy of artificial intelligence.)
  9. This list of intelligent traits is based on the topics covered by major AI textbooks, including: 脚本错误:没有“Footnotes”这个模块。, 脚本错误:没有“Footnotes”这个模块。, 脚本错误:没有“Footnotes”这个模块。 and 脚本错误:没有“Footnotes”这个模块。.
  10. Pfeifer, R. and Bongard J. C., How the body shapes the way we think: a new view of intelligence (The MIT Press, 2007).
  11. White, R. W. (1959). "Motivation reconsidered: The concept of competence". Psychological Review. 66 (5): 297–333. doi:10.1037/h0040934. PMID 13844397.
  12. 脚本错误:没有“Footnotes”这个模块。
  13. deCharms, R. (1968). Personal causation. New York: Academic Press.
  14. Muehlhauser, Luke (11 August 2013). "What is AGI?". Machine Intelligence Research Institute. Archived from the original on 25 April 2014. Retrieved 1 May 2014.
  15. "What is Artificial General Intelligence (AGI)? | 4 Tests For Ensuring Artificial General Intelligence". Talky Blog (in English). 13 July 2019. Archived from the original on 17 July 2019. Retrieved 17 July 2019.
  16. Shapiro, Stuart C. (1992). Artificial Intelligence -{zh-cn:互联网档案馆; zh-tw:網際網路檔案館; zh-hk:互聯網檔案館;}-存檔,存档日期1 February 2016. In Stuart C. Shapiro (Ed.), Encyclopedia of Artificial Intelligence (Second Edition, pp. 54–57). New York: John Wiley. (Section 4 is on "AI-Complete Tasks".)
  17. Roman V. Yampolskiy. Turing Test as a Defining Feature of AI-Completeness. In Artificial Intelligence, Evolutionary Computation and Metaheuristics (AIECM) --In the footsteps of Alan Turing. Xin-She Yang (Ed.). pp. 3–17. (Chapter 1). Springer, London. 2013. http://cecs.louisville.edu/ry/TuringTestasaDefiningFeature04270003.pdf -{zh-cn:互联网档案馆; zh-tw:網際網路檔案館; zh-hk:互聯網檔案館;}-存檔,存档日期22 May 2013.
  18. Luis von Ahn, Manuel Blum, Nicholas Hopper, and John Langford. CAPTCHA: Using Hard AI Problems for Security -{zh-cn:互联网档案馆; zh-tw:網際網路檔案館; zh-hk:互聯網檔案館;}-存檔,存档日期4 March 2016.. In Proceedings of Eurocrypt, Vol. 2656 (2003), pp. 294–311.
  19. Bergmair, Richard (7 January 2006). "Natural Language Steganography and an "AI-complete" Security Primitive". CiteSeerX 10.1.1.105.129. Cite journal requires |journal= (help) (unpublished?)
  20. 脚本错误:没有“Footnotes”这个模块。
  21. 脚本错误:没有“Footnotes”这个模块。 quoted in 脚本错误:没有“Footnotes”这个模块。
  22. "Scientist on the Set: An Interview with Marvin Minsky". Archived from the original on 16 July 2012. Retrieved 5 April 2008.
  23. Marvin Minsky to 脚本错误:没有“Footnotes”这个模块。, quoted in 脚本错误:没有“Footnotes”这个模块。.
  24. The Lighthill report specifically criticized AI's "grandiose objectives" and led the dismantling of AI research in England. (脚本错误:没有“Footnotes”这个模块。; 脚本错误:没有“Footnotes”这个模块。) In the U.S., DARPA became determined to fund only "mission-oriented direct research, rather than basic undirected research". See 模板:Harv under "Shift to Applied Research Increases Investment". See also 模板:Harv and 模板:Harv
  25. 脚本错误:没有“Footnotes”这个模块。, 脚本错误:没有“Footnotes”这个模块。 and see also 脚本错误:没有“Footnotes”这个模块。
  26. 脚本错误:没有“Footnotes”这个模块。; 脚本错误:没有“Footnotes”这个模块。; 脚本错误:没有“Footnotes”这个模块。
  27. 脚本错误:没有“Footnotes”这个模块。
  28. As AI founder John McCarthy writes "it would be a great relief to the rest of the workers in AI if the inventors of new general formalisms would express their hopes in a more guarded form than has sometimes been the case." McCarthy, John (2000). "Reply to Lighthill". Stanford University. Archived from the original on 30 September 2008. Retrieved 29 September 2007.
  29. "At its low point, some computer scientists and software engineers avoided the term artificial intelligence for fear of being viewed as wild-eyed dreamers."Markoff, John (14 October 2005). "Behind Artificial Intelligence, a Squadron of Bright Real People". The New York Times.
  30. 脚本错误:没有“Footnotes”这个模块。
  31. "Trends in the Emerging Tech Hype Cycle". Gartner Reports. Archived from the original on 22 May 2019. Retrieved 7 May 2019.
  32. 脚本错误:没有“Footnotes”这个模块。
  33. Harnad, S (1990). "The Symbol Grounding Problem". Physica D. 42 (1–3): 335–346. arXiv:cs/9906002. Bibcode:1990PhyD...42..335H. doi:10.1016/0167-2789(90)90087-6.
  34. 脚本错误:没有“Footnotes”这个模块。
  35. "Who coined the term "AGI"? » goertzel.org" (in English). Archived from the original on 28 December 2018. Retrieved 28 December 2018., via Life 3.0: 'The term "AGI" was popularized by... Shane Legg, Mark Gubrud and Ben Goertzel'
  36. 脚本错误:没有“Footnotes”这个模块。. See also 脚本错误:没有“Footnotes”这个模块。 with an up-to-date summary and lots of links.
  37. https://goertzel.org/AGI_Summer_School_2009.htm
  38. http://fmi-plovdiv.org/index.jsp?id=1054&ln=1
  39. http://fmi.uni-plovdiv.bg/index.jsp?id=1139&ln=1
  40. 40.0 40.1 40.2 模板:Harv or see Advanced Human Intelligence -{zh-cn:互联网档案馆; zh-tw:網際網路檔案館; zh-hk:互聯網檔案館;}-存檔,存档日期30 June 2011. where he defines strong AI as "machine intelligence with the full range of human intelligence."
  41. Markoff, John (27 November 2016). "When A.I. Matures, It May Call Jürgen Schmidhuber 'Dad'". The New York Times. Archived from the original on 26 December 2017. Retrieved 26 December 2017.
  42. James Barrat (2013). "Chapter 11: A Hard Takeoff". Our Final Invention: Artificial Intelligence and the End of the Human Era (First ed.). New York: St. Martin's Press. ISBN 9780312622374. 
  43. "About the Machine Intelligence Research Institute". Machine Intelligence Research Institute. Archived from the original on 21 January 2018. Retrieved 26 December 2017.
  44. "About OpenAI". OpenAI (in English). Archived from the original on 22 December 2017. Retrieved 26 December 2017.
  45. Theil, Stefan. "Trouble in Mind". Scientific American (in English). pp. 36–42. Bibcode:2015SciAm.313d..36T. doi:10.1038/scientificamerican1015-36. Archived from the original on 9 November 2017. Retrieved 26 December 2017.
  46. Liu, Feng; Shi, Yong; Liu, Ying (2017). "Intelligence Quotient and Intelligence Grade of Artificial Intelligence". Annals of Data Science. 4 (2): 179–191. arXiv:1709.10242. Bibcode:2017arXiv170910242L. doi:10.1007/s40745-017-0109-0.
  47. "Google-KI doppelt so schlau wie Siri". Archived from the original on 3 January 2019. Retrieved 2 January 2019.
  48. Lawler, Richard (November 13, 2019). "John Carmack takes a step back at Oculus to work on human-like AI". Engadget. Retrieved April 4, 2020.
  49. 实现通用智能行为的一种被广泛的讨论方法是全脑模拟。一个低层次的大脑模型是通过扫描和绘制生物大脑的详细情况,并将其状态复制到计算机系统或其他计算设备中来建立的。计算机运行的模拟模型无比忠实于原始模型,以至于它的行为在本质,或者所有实际目的上与原始大脑相同,难以区分。 脚本错误:没有“Footnotes”这个模块。. "The basic idea is to take a particular brain, scan its structure in detail, and construct a software model of it that is so faithful to the original that, when run on appropriate hardware, it will behave in essentially the same way as the original brain."
  50. In "Mind Children" 脚本错误:没有“Footnotes”这个模块。 1015 cps is used. More recently, in 1997, <"Archived copy". Archived from the original on 15 June 2006. Retrieved 23 June 2006.CS1 maint: archived copy as title (link)> Moravec argued for 108 MIPS which would roughly correspond to 1014 cps. Moravec talks in terms of MIPS, not "cps", which is a non-standard term Kurzweil introduced.
  51. Swaminathan, Nikhil (Jan–Feb 2011). "Glia—the other brain cells". Discover. Archived from the original on 8 February 2014. Retrieved 24 January 2014.
  52. Izhikevich, Eugene M.; Edelman, Gerald M. (4 March 2008). "Large-scale model of mammalian thalamocortical systems" (PDF). PNAS. 105 (9): 3593–3598. Bibcode:2008PNAS..105.3593I. doi:10.1073/pnas.0712231105. PMC 2265160. PMID 18292226. Archived from the original (PDF) on 12 June 2009. Retrieved 23 June 2015.
  53. "Project Milestones". Blue Brain. Retrieved 11 August 2008.
  54. "Artificial brain '10 years away' 2009 BBC news". 22 July 2009. Archived from the original on 26 July 2017. Retrieved 25 July 2009.
  55. University of Calgary news -{zh-cn:互联网档案馆; zh-tw:網際網路檔案館; zh-hk:互聯網檔案館;}-存檔,存档日期18 August 2009., NBC News news -{zh-cn:互联网档案馆; zh-tw:網際網路檔案館; zh-hk:互聯網檔案館;}-存檔,存档日期4 July 2017.
  56. "Archived copy". Archived from the original on 15 June 2006. Retrieved 23 June 2006.CS1 maint: archived copy as title (link)
  57. 脚本错误:没有“Footnotes”这个模块。. A wide range of views in current research, all of which require grounding to some degree
  58. "some links to bee brain studies". Archived from the original on 5 October 2011. Retrieved 30 March 2010.
  59. Yekutieli, Y; Sagiv-Zohar, R; Aharonov, R; Engel, Y; Hochner, B; Flash, T (August 2005). "Dynamic model of the octopus arm. I. Biomechanics of the octopus reaching movement". J. Neurophysiol. 94 (2): 1443–58. doi:10.1152/jn.00684.2004. PMID 15829594.
  60. 脚本错误:没有“Footnotes”这个模块。
  61. "nervous system, human." Encyclopædia Britannica. 9 January 2007
  62. 脚本错误:没有“Footnotes”这个模块。
  63. As defined in a standard AI textbook: "The assertion that machines could possibly act intelligently (or, perhaps better, act as if they were intelligent) is called the 'weak AI' hypothesis by philosophers, and the assertion that machines that do so are actually thinking (as opposed to simulating thinking) is called the 'strong AI' hypothesis." 模板:Harv
  64. For example:
  65. Note that consciousness is difficult to define. A popular definition, due to Thomas Nagel, is that it "feels like" something to be conscious. If we are not conscious, then it doesn't feel like anything. Nagel uses the example of a bat: we can sensibly ask "what does it feel like to be a bat?" However, we are unlikely to ask "what does it feel like to be a toaster?" Nagel concludes that a bat appears to be conscious (i.e. has consciousness) but a toaster does not. See 模板:Harv
  66. Sotala, Kaj; Yampolskiy, Roman V (2014-12-19). "Responses to catastrophic AGI risk: a survey". Physica Scripta. 90 (1): 8. doi:10.1088/0031-8949/90/1/018001. ISSN 0031-8949.
  67. Joy, Bill (April 2000). "Why the future doesn't need us". Wired.
  68. Kaplan, Andreas; Haenlein, Michael (2019). "Kaplan Andreas and Haelein Michael (2019) Siri, Siri, in my hand: Who's the fairest in the land? On the interpretations, illustrations, and implications of artificial intelligence". Business Horizons. 62: 15–25. doi:10.1016/j.bushor.2018.08.004.
  69. europarl.europa.eu: How artificial intelligence works, "Concluding remarks: Today's AI is powerful and useful, but remains far from speculated AGI or ASI.", European Parliamentary Research Service, retrieved March 3, 2020
  70. itu.int: Beyond Mad?: The Race For Artificial General Intelligence, "AGI represents a level of power that remains firmly in the realm of speculative fiction as on date." February 2, 2018, retrieved March 3, 2020
  71. Allen, Paul. "The Singularity Isn't Near". MIT Technology Review. Retrieved 17 September 2014.
  72. Winfield, Alan. "Artificial intelligence will not turn into a Frankenstein's monster". The Guardian. Archived from the original on 17 September 2014. Retrieved 17 September 2014.
  73. Raffi Khatchadourian (23 November 2015). "The Doomsday Invention: Will artificial intelligence bring us utopia or destruction?". The New Yorker. Archived from the original on 28 January 2016. Retrieved 7 February 2016.
  74. Müller, V. C., & Bostrom, N. (2016). Future progress in artificial intelligence: A survey of expert opinion. In Fundamental issues of artificial intelligence (pp. 555–572). Springer, Cham.
  75. Rawlinson, Kevin. "Microsoft's Bill Gates insists AI is a threat". BBC News. Retrieved 30 January 2015.
  76. Bostrom, Nick (2014). Superintelligence: Paths, Dangers, Strategies (First ed.). ISBN 978-0199678112. 
  77. Kaj Sotala; Roman Yampolskiy (19 December 2014). "Responses to catastrophic AGI risk: a survey". Physica Scripta. 90 (1).
  78. "But What Would the End of Humanity Mean for Me?". The Atlantic. 9 May 2014. Retrieved 12 December 2015.
  79. "Tech Luminaries Address Singularity". IEEE Spectrum: Technology, Engineering, and Science News (in English) (SPECIAL REPORT: THE SINGULARITY). 1 June 2008. Retrieved 8 April 2020.
  80. Shermer, Michael (1 March 2017). "Apocalypse AI". Scientific American (in English). p. 77. Bibcode:2017SciAm.316c..77S. doi:10.1038/scientificamerican0317-77. Retrieved 27 November 2017.


References 参考文献

• Stages of Artificial Intelligence"Computer Science" on 2nd April, 2020.

• Stages of Artificial Intelligence"Computer Science" on 2nd April, 2020.

•2020年4月2日,《人工智能的阶段》[ https://www.computerscience0.xyz/2020/04/ai-Artificial-Intelligence-Stages-of.html 计算机科学]。

  • Omohundro, Steve (2008), The Nature of Self-Improving Artificial Intelligence, presented and distributed at the 2007 Singularity Summit, San Francisco, CA.
  • Sandberg, Anders; Boström, Nick (2008), Whole Brain Emulation: A Roadmap (PDF), Technical Report #2008‐3, Future of Humanity Institute, Oxford University, retrieved 5 April 2009
  • Berglas

最后一个贝格拉斯, Anthony

首先是安东尼 (2008

2008年), [http://berglas.org/Articles/AIKillGrandchildren/AIKillGrandchildren.html

Http://berglas.org/articles/aikillgrandchildren/aikillgrandchildren.html Artificial Intelligence will Kill our Grandchildren 人工智能会杀死我们的孙子] Check |url= value (help) line feed character in |last= at position 8 (help); line feed character in |title= at position 52 (help); line feed character in |first= at position 8 (help); line feed character in |year= at position 5 (help); line feed character in |url= at position 73 (help); Check date values in: |year= (help)

}}

}}

  • Darrach, Brad (20 November 1970), "Meet Shakey, the First Electronic Person", Life Magazine, pp. 58–68.
  • Gelernter, David

最后的盖兰特 (2010

2010年), Dream-logic, the Internet and Artificial Thought 梦的逻辑,互联网和人工思维, retrieved 25 July 2010

Http://www.edge.org/3rd_culture/gelernter10.1/gelernter10.1_index.html : 2010年7月25日 Text " 第一个大卫

" ignored (help); line feed character in |year= at position 5 (help); line feed character in |title= at position 49 (help); line feed character in |first= at position 6 (help); line feed character in |accessdate= at position 13 (help); Check date values in: |accessdate=, |year= (help)

}}

}}

  • Goertzel, Ben; Pennachin, Cassio, eds. (2006

2006年), Artificial General Intelligence 人工通用智能 (PDF), Springer

出版商斯普林格, ISBN [[Special:BookSources/978-3-540-23733-4

[国际标准图书馆编号978-3-540-23733-4]|978-3-540-23733-4 [国际标准图书馆编号978-3-540-23733-4]]] Check |isbn= value: invalid character (help), archived from [http://people.inf.elte.hu/csizsekp/ai/books/artificial-general-intelligence-cognitive-technologies.9783540237334.27156.pdf

Http://people.inf.elte.hu/csizsekp/ai/books/artificial-general-intelligence-cognitive-technologies.9783540237334.27156.pdf the original] Check |url= value (help) (PDF) on 20 March 2013 Text " authorlink Ben Goertzel

" ignored (help); Text " 编辑2-last Pennachin " ignored (help); Text " 档案-url https://web.archive.org/web/20130320184603/http://people.inf.elte.hu/csizsekp/ai/books/artificial-general-intelligence-cognitive-technologies.9783540237334.27156.pdf

" ignored (help); Text " 档案-日期2013年3月20日

" ignored (help); Text " 编辑2-first Cassio

" ignored (help); line feed character in |isbn= at position 18 (help); line feed character in |publisher= at position 10 (help); line feed character in |year= at position 5 (help); line feed character in |title= at position 32 (help); line feed character in |url= at position 123 (help); line feed character in |authorlink= at position 13 (help); Check date values in: |year= (help)

}}

}}

  • [[Ben Goertzel

作者: Ben Goertzel|Goertzel, Ben]]; Wang, Pei

2 Wang (2006

2006年), [http://sites.google.com/site/narswang/publications/wang-goertzel.AGI_Aspects.pdf?attredirects=1

Http://sites.google.com/site/narswang/publications/wang-goertzel.agi_aspects.pdf?attredirects=1 Introduction: Aspects of Artificial General Intelligence] Check |url= value (help) (PDF) Text " first2 Pei

" ignored (help); Text " 题目简介: 人工通用智能的方方面面

" ignored (help); line feed character in |year= at position 5 (help); line feed character in |url= at position 96 (help); line feed character in |authorlink= at position 13 (help); line feed character in |first2= at position 4 (help); Check date values in: |year= (help)

}}

}}

  • Lighthill, Professor Sir James (1973), "Artificial Intelligence: A General Survey", Artificial Intelligence: a paper symposium, Science Research Council
  • Newell, Allen; Simon, H. A. (1963), "GPS: A Program that Simulates Human Thought", in Feigenbaum, E.A.; Feldman, J. (eds.), Computers and Thought, New York: McGraw-Hill
  • Simon, H. A. (1965), The Shape of Automation for Men and Management, New York: Harper & Row
  • Sutherland, J.G. (1990), "Holographic Model of Memory, Learning, and Expression", International Journal of Neural Systems, 1–3: 256–267.
  • de Vega, Manuel; Glenberg, Arthur

2- 最后的格伦伯格2- 第一个亚瑟; Graesser, Arthur, eds. (2008

2008年), Symbols and Embodiment: Debates on meaning and cognition 标题符号与具体化: 关于意义与认知的争论, Oxford University Press

牛津大学出版社, ISBN [[Special:BookSources/978-0-19-921727-4

[国际标准图书编号978-0-19-921727-4]|978-0-19-921727-4 [国际标准图书编号978-0-19-921727-4]]] Check |isbn= value: invalid character (help) Text " 编辑1-last de Vega " ignored (help); Text " 编辑1-first Manuel

" ignored (help); Text " 编辑3-first Arthur

" ignored (help); Text " 编辑3-last Graesser " ignored (help); line feed character in |editor2-first= at position 7 (help); line feed character in |isbn= at position 18 (help); line feed character in |publisher= at position 24 (help); line feed character in |year= at position 5 (help); line feed character in |title= at position 57 (help); Check date values in: |year= (help)

}}

}}


External links 拓展链接


模板:Existential risk from artificial intelligence

Category:Hypothetical technology

类别: 假设技术

Category:Artificial intelligence

类别: 人工智能

Category:Computational neuroscience

类别: 计算神经科学


fr:Intelligence artificielle#Intelligence artificielle forte

fr:Intelligence artificielle#Intelligence artificielle forte

智力人工 # 智力人工强项


This page was moved from wikipedia:en:Artificial general intelligence. Its edit history can be viewed at 通用人工智能/edithistory