- 最后登录
- 2010-6-24
- 在线时间
- 63 小时
- 寄托币
- 564
- 声望
- 14
- 注册时间
- 2009-9-1
- 阅读权限
- 25
- 帖子
- 17
- 精华
- 0
- 积分
- 353
- UID
- 2691719

- 声望
- 14
- 寄托币
- 564
- 注册时间
- 2009-9-1
- 精华
- 0
- 帖子
- 17
|
本帖最后由 windLee 于 2009-9-3 10:32 编辑
--------------------------------------------
第一段,质疑机器是否有超越人类的可能性,得到答案:单纯的工具是不行的,但是,向着超越人类的目标去设计的机器则有可能,因为它们不只是工具而已。
第二段:讲述纯粹的工具与人工智能的区别,例子是,工具如汽车,计算器等仅仅是对人类力量和智慧的加强而已;而独立思考的机器如IBM的下棋机则是另外一回事。然后说出为什么人们有一种人工智能都很弱智的错误印象及其原因。
第三段:讲述人工智能的新研究方向,以及基于这样的基础,带来的它们自我学习和思考,以及超越人类的可能性,此外,已经在某些方面超越了人类的事实。
第四段:补充说明现在对于机器研发的一种良好思路,给我们的工具配备一些独立思考的能力,使得它们不单单只是dumb tools(哑巴工具),而更能够借助自己的判断力来辅助人类, 从而,智能机器和人类的组合成为既超越人类,又超越工具的存在。
最后一段:总结,并不是所有的机器都只是工具,工具是没有智能的人类延伸,而专门设计的智能机器,总有一天会超越人类。
-------------------------------------------------------------
"The human mind will always be superior to machines because machines are only tools of human minds."
Once I asked my mentor a question, that is, is it possible for a man to create a machine which is cleverer than him? And I presented my hypothesis according to my sensation that none could build such a machine. However, after the invention of the famous IBM chess robot "deep blue" and its near triumph over the worlds' best human chess player, the answer to the question seems shifted a bit, as there is no way those programmers for "deep blue" could compare to the best chess player in playing chess. As a result, even though the "deep blue" lost, it is quite possible for its chess skill to exceed its inventor's. Thus, if the machines are only designed as dumb tools maybe they will not be superior to human minds, but if they are created under certain rules and they can learn by themselves for how to use those rules, their fast calculating ability might give them the chance to take advantage over humans.
The identification of machines has changed through ages, from tools to independent intelligent systems. In the first few decades of their invention, it seems they are merely mindless tools which are used to enhance human strength, mostly to augment the physical power, or to increase the calculating speed. An apposite icon is the car which could take people around fast while carry many weights, another instance is a scientific calculator which could do many mathematic jobs in a flash. No matter what functions they have, they are auxiliary tools which are mindless, not having the chance to be compared to human minds. Nevertheless, there are machines designed to "think" in some logic ways. They are called the artificial intelligence. The "deep blue" is a significant example as it's not a chess tool, it plays chess by itself but not under human manipulation or guidance. In this aspect, these machines are more like individuals other than apparatus for humans. Indeed, there is a misunderstanding about the artificial intelligence. As there are certain limitations in computer science and the studies of human brain, artificial intelligence is quite fettered by these low technology, thus most artificial intelligences only equal to a "3 years-old child", and the media reports make the public believe it is extremely hard to make this "baby" grow even for 1 age.
In fact, as the computer technology has developed greatly in recent years, there are brand new directions to develop the artificial intelligence. The "deep blue" is a sign of the AI to be possibly superior to human minds in specific fields. And new researches began to focus and pursue multicore system, turned out to a brilliant way of expanding computing capabilies and functions, rather than the single but fast calculating unite. This theory has been brought into the design of AI, where scientists use groups of simple-functioned self-learning mini robots to communicate and form an intelligent community(e.g. the nerve network), while the industries of micro integrated circuit and system on chip ensured an increasingly rapid center unit of these robots. Thus, if the autonomy is granted by these mini communities, they gain the ability to learn and evolve, which guarantees them the possibility to be more intelligent than single human minds.
While the development of independent AI systems goes on, the best way to create a smart and powerful combination at this era is to make the tools for human minds cleverer and easier to use. For example, the driving system of a car is always a highly-marked research project which requires both intelligence and experience on steering and situation analysis. Humans make a lot mistakes which cause accidents in driving, proving themselves incapable of complete safe drive. Thus the machines are required to help make quick judgment when emergency happens as they are competent for fast calculating. Research on this project has bloomed and achieved many products. They can help people drive more stable on slipping grounds, snow-covered roads and narrow corners, even a rookie could drive much safer now. Are they able to drive alone and become not only steering tools? Definitely, the answer is absolute. Not need to mention they already drive better than us in the games, there also exists a system called auto-navigation in reality now, which ensures the blind people to drive a car! This machine could not be called a tool anymore, as it totally acts independently.
In conclusion, our civilization has never ended our development on artificial intelligence, and these "machines" are becoming gradually solo players with its own "thoughts", but not a tool under the control of human minds. A tool can never be smart and more advanced than its user but a machine designed to have intelligence might be able to do that in the future.
-------------------------------
--------------------------------
这一篇是用笔记本敲的,确实比台式键盘快多了。不过我比较罗嗦.. 感觉有很多废话.. 自己语法也不太擅长,计时作文的话有效词汇也变得空白.
总之,拜托大家帮忙看一看,我知道有付出才能得到帮助, 不过考试在即不免有些急切。 今后也会努力但愿能给大家帮上忙。 |
|