Recent advances in artificial intelligence have sparked fears that the machines may one day take over
What distinguishes human beings from AI is consciousness of the kind that leads to independent and creative thinking
Author: Weijian Shan
Published: 1:00am, 29 Mar, 2023
South China Moring Post
A visitor takes a picture with his mobile phone of an image designed with artificial intelligence by Berlin-based digital creator Julian van Dieken inspired by Johannes Vermeer’s painting “Girl with a Pearl Earring” at the Mauritshuis museum in The Hague on March 9. Photo: AFP
Recent advances in artificial intelligence raise the question of whether it will eventually replace and surpass human intelligence.
A decade ago physicist Stephen Hawking said he feared thinking machines would evolve faster than humans, bound as we are by our biological limits, and eventually take on a life of their own, potentially spelling the end of the human race.
Can machines become intelligent and powerful enough to turn against their makers, as they did the film The Matrix?
The answer is no, because machines cannot acquire the “consciousness” – the awareness of one’s own existence, thoughts, feelings, sensations and surroundings – that would inspire such a revolt.
Consciousness can be seen as a gift from God, granted only to humans. Scientifically speaking, consciousness evolved from the primordial need to survive and reproduce, which gave rise to the “desire” and “intentions” that drive behaviour and set objectives. An AI machine will not evolve to set objectives on its own, because without these primordial needs it cannot acquire consciousness or free will.
AI is able to discover patterns in huge quantities of data far better than humans. But it can only do so for an objective defined by a human being. A computer can beat a human in a game of chess whose objective and rules have already been set by a human being, but it cannot change the objective or the rules.
Why does this matter? Without consciousness, desire, intent and the ability to set goals, AI, no matter how powerful, will be incapable of the independent, creative and critical thinking necessary to come up with new ideas – the ability to theorise, hypothesise, postulate or imagine with limited or no information, before such ideas can be tested or proven by data, observations or experiments. Imagination is only human.
Imagine there had been an AI that incorporated all the knowledge available at the time of Christopher Columbus. Would it have been able to introduce Europeans to the “New World”? No, because there was no information about the New World or about a westward route to India. Finding that route required having a new idea, a theory, a purpose, free will and a desire – which can only come from a human mind.
Albert Einstein pondered the theories of relativity through visualised thought experiments, long before they were proven. At the age of 16, he imagined himself chasing a beam of light at the same speed and hit upon the paradox that whereas he would observe the beam to be at a standstill – just as the seats in an aeroplane do not move relative to its passengers – someone on the Earth would see it travelling at the velocity of light.
Of course, he needed all the contemporary knowledge of physics and mathematics, which he acquired later in life, to develop his theories. But his ability to imagine things from nothing can only be a human endeavour.
Consider another human trait – deception. Humans lie, sometimes for good purposes, but often for bad ones. “Operation Mincemeat” was a deception the British cooked up during World War II to misdirect Hitler’s forces away from the Allied invasion of Sicily. Bernard Madoff cheated his investors and enriched himself with a Ponzi scheme. Can you imagine AI lying or cheating unless a human being directs it to do so?
AI cannot be a philosopher or spiritual leader because machines cannot be made capable of abstract thinking or of inspiration. Without these qualities, there would be no Aristotle, Buddha or Martin Luther, all of whom produced profound insights into the nature of the universe and the human condition.
Nor can AI conduct business negotiations whose outcome cannot be predefined. My new book, Money Machine, tells the story of multiparty negotiations for a history-making deal whose outcome could have been any of the infinite possibilities, none of which was optimal for any party.
It is consciousness that ultimately defines the boundaries between human and artificial intelligence. Consciousness leads to independent and creative thinking that produces novel ideas, such as those of Isaac Newton, Plato and Confucius.
This is not to downplay the significance of AI. It will increasingly be a powerful instrument that help humans achieve their objectives. But it will remain just that, as have all tools since the dawn of human consciousness.
Weijian Shan is the author of Out of the Gobi, Money Games, and Money Machine, a trustee of the British Museum and executive chairman of PAG
人工智能永远无法比拟人类智慧
单伟建
近来人工智能的快速发展产生了一个问题，即电脑是否终将取代人脑，甚至超越人类的智慧。
理论物理学家史蒂芬·霍金十年前说过，会思考的机器会比人类智力进化的速度更快，因为人类智力的进化受限于生物本质的局限，所以机器最终会超越人的智力，“那将带来人类末日”。
1999年上映的大片《黑客帝国》就是一个机器奴役了人类的科幻故事。人工智能的发展可以使这类幻想成真吗？
答案是否定的。人工智能永远无法取代人类的智慧，更不可能奴役人类。这是因为再聪明的机器也无法产生“意识”，没有意识，就没有意愿，也就不能自行其是，更不可能造反。
什么是意识？意识是感知自己的存在、自己的思想、记忆、感受以及感知周围的存在。
意识从何而来？来自于大脑，大脑进化到一定程度，就产生了意识。大脑之所以进化，为的是适者生存，归根结底是生存和繁衍的需要。
人工智能可以越来越聪明，可以无限进步，但永远没有生存和繁衍需要，永远不会产生欲望和意愿，因此也无法产生意识。
电脑算力无限提高使人工智能可以即时处理海量信息，从中找出模式、规律，解决问题。电脑已经学会下棋，击败任何人类高手。但是它能够击败人类的任何游戏的规则都是人类制定的，由于没有意识，电脑不可能自主改变游戏规则。只有有意识的人，才会有意愿改变游戏规则，改变要达到的目的。
有了意识才能有想象力。想象是抽象思维，思考实际不存在的东西。借助想象力，人类才能产生新思想、新概念、新理论、新假设，乃至新价值观。人类的发现和发明，往往是理论或假设在先，而求证在后。胡适之所倡导的“大胆假设，小心求证”就是发现的一般途径。
假设远在哥伦布时代就有了人工智能，其涵括了到那时为止所有已知的知识和信息，这个聪明的机器人能够将新大陆介绍给欧洲人吗？不可能。因为当时对于新大陆的存在一无所知。机器所储存的信息可以涵括全世界的书籍，人脑则望尘莫及，但发现新大陆需要有意识，有欲望，有想象力，有理论，有理想，有信念，有意志，所有这些只能来自人的大脑。
爱因斯坦发现相对论的重要途径是以形象思维的方式做“思想实验”。他在自传中说，16岁的时候就想象自己以光速追逐一束光线，就此发现了一个悖论，即与一束光同行的他，应该看到的是一束静止的光线（正如飞机上的座椅相对于乘客是静止的），但地球上的观察者看到的同一束光却是以光速移动。
当然，从追逐光束的想象到真正发展出相对论还有待于他学习和掌握了当代的物理知识和相当的数学基础之后，而相对论的证明则更在若干年之后。重要的是，抽象思维和凭空想象是人类才具有的能力。
人类还有一种自主的能力是人工智能无法具备的，这就是欺骗。人类会撒谎。谎言有善意的（譬如告诉得了绝症的病人他很快就会康复），多为居心不良。兵不厌诈，二次世界大战期间，英国的情报机构执行了一个“碎肉行动”，编织了一套弥天大谎，将希特勒的部队骗离西西里岛，从而保证了盟军登陆成功。用谎言欺诈的案子比比皆是，典型的例子就是在2008年败露的伯纳德·麦道夫（Bernard Madoff），他用庞氏骗局的手段诈骗投资者以自肥。若无人指使，人工智能机不可能撒谎。
人工智能不可能成为哲学家或精神领袖。如果没有抽象思维和想象的能力，没有灵感的驱动，就不会有亚里士多德、孔夫子或者释迦牟尼。这些达到人类智慧巅峰的圣贤是通过观察和感受人类的幸福和痛苦而产生了对于自然界和人性的深刻认识。
人工智能可以取代政客做决策吗？很遗憾，永远不可能。因为人类有情感、理想、信念和主观能动性，人的决策并非总是被民意测验或统计模型所左右。同样，谈判也无法让人工智能越俎代庖，因为与下棋或玩游戏不同，谈判的结果并非简单的输赢，可有无数可能性，而无一最优。
人类智慧和人工智能的分水岭就是自主意识。意识使人类可以独立、创造性思考，摈弃已有认知而创造全新理论和思想体系，正因如此，才有孔夫子、柏拉图、牛顿、爱因斯坦等古往今来的至圣先师。
这并非低估人工智能划时代颠覆性的重大意义。人工智能会不断强大，解决很多以往人类因为大脑的局限而解决不了的问题，帮助人类实现理想和目的。但无论多么强大，它仍然为我所用，仍然是人类的工具，就像从人类获得自我意识之初到今天所创造的一切工具一样。
单伟建是《走出戈壁》（Out of the Gobi）、《金钱博弈》（Money Games）以及《Money Machine》的作者、大英博物馆董事、太盟投资集团执行董事长
2023年3月28日於日内瓦
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
