ai 项目经理 要了解
Simple and useful guide!
简单实用的指南!
Artificial Intelligence (AI) has taken over the world by storm in recent years. Artificial Intelligence is one of the most popular terms of Computer Science. Artificial Intelligence, also known as Machine Intelligence, is a simulation of human intelligence in machines that are designed to think like humans and mimic their actions. AI makes sure that the tasks performed by the machines are very similar to that of humans, such as driving a car or asking for directions. Applications of AI are powered by some of the advanced tools and technologies like Machine Learning, Deep Learning, Natural Language Processing, and Robotics.
近年来,人工智能(AI)已席卷全球。 人工智能是计算机科学中最受欢迎的术语之一。 人工智能(也称为机器智能)是在旨在像人类一样思考并模仿其行为的机器中模拟人类智能。 AI确保机器执行的任务与人类的任务非常相似,例如驾驶汽车或询问方向。 AI的应用由机器学习,深度学习,自然语言处理和机器人等一些高级工具和技术提供支持。
When people hear about Artificial Intelligence, the first thing which occurs to their mind is Robots. That is because of the sci-fi movies, books, games, and those fascinating predictions about the world being ruled by robots, flying cars, and transformers. As we are going deep in the AI world, one thing which makes us wonder is that why we still don’t live in a world as depicted in the movies like The Matrix or Blade Runner 2049? The truth is that, despite being surrounded by such movies or books, few of us have learned to use the term “AI” under the right circumstances. Expecting too much from it can hurt our expectations and assumptions about how the future looks like.
当人们听说人工智能时,他们想到的第一件事就是机器人。 这是因为科幻电影,书籍,游戏以及关于机器人,飞行汽车和变压器统治世界的那些有趣的预测。 随着我们深入AI世界,令我们感到奇怪的是,为什么我们仍然不生活在《黑客帝国》或《银翼杀手2049》等电影所描绘的世界中? 事实是,尽管被这类电影或书籍所包围,但我们很少有人学会在适当的情况下使用“人工智能”一词。 对它的期望过高会损害我们对未来前景的期望和假设。
Artificial Intelligence broadly works on the principle that human intelligence can be easily mimicked by the machines to perform human-like tasks, from simplest to hardest. The important goals of AI are problem-solving, decision-making, perception, learning, natural language processing, planning, and reasoning. With all the advancements in the technology, the prototypes that defined AI earlier become outdated. AI researches and experts around the world are working together endlessly to build a future driven by the rise of this new technology. What seems impossible today, becomes possible tomorrow. As we know, AI is impacting all the industries and changing the world exponentially, so it becomes crucial that we think seriously about this new field of computer science. To live in the future surrounded by flying cars and robots, we must understand the implication of AI on our society, and to know where we stand today, we first need to know the types of AI –
人工智能广泛地基于这样的原理,即机器可以很容易地模仿人类的智能,以执行从最简单到最困难的类似人类的任务。 人工智能的重要目标是解决问题,决策,感知,学习,自然语言处理,计划和推理。 随着技术的所有进步,定义AI的原型已经过时了。 世界各地的AI研究人员和专家正在不懈努力,以建立在这项新技术兴起的推动下的未来。 今天似乎不可能的事,明天变得可能。 众所周知,人工智能正影响着所有行业,并且正在以指数方式改变世界,因此,认真考虑计算机科学这一新领域至关重要。 为了生活在被飞行的汽车和机器人包围的未来中,我们必须了解AI对我们社会的影响,并且要了解我们今天的地位,我们首先需要了解AI的类型-
Artificial Intelligence consists of three main types –
人工智能包括三种主要类型:
- Artificial Narrow Intelligence or Weak AI 人工智能或弱人工智能
- Artificial General Intelligence or Strong AI 人工智能或强人工智能
- Artificial Super Intelligence or Super AI 人工超级智能或超级AI
1.人工智能或弱人工智能 (1. Artificial Narrow Intelligence or Weak AI)
Artificial Narrow Intelligence, or Weak AI, or Narrow AI represents the AI technology that exists in today’s world. Artificial Narrow Intelligence refers to a computer’s ability to perform narrow or single tasks extremely well and sometimes maybe even better than humans. ANI is what we’ve been able to accomplish and implement successfully so far. ANI performs the given tasks in real-time while pulling information from a large dataset. These machines therefore can’t do anything more than what they’re programmed to do, thus having very limited ability to do things efficiently.
人工窄智能或弱AI或窄AI代表了当今世界上存在的AI技术。 人工窄带智能是指计算机能够非常好地执行狭窄或单个任务的能力,有时甚至比人类更好。 到目前为止,ANI是我们能够成功完成和实施的。 ANI实时执行给定的任务,同时从大型数据集中提取信息。 因此,这些机器只能完成其编程的工作,因此有效执行工作的能力非常有限。
There are many applications of Narrow AI around us today and becoming constantly common as Machine Learning. Devices like Siri, Google Assistance, Alexa, Cortana all are great examples of ANI. Some other examples include –
如今,我们周围有Narrow AI的许多应用,并且随着机器学习而变得越来越普遍。 诸如Siri,Google Assistance,Alexa,Cortana之类的设备都是ANI的绝佳示例。 其他一些例子包括–
- Self-driving cars 自动驾驶汽车
- Credit card fraud detection 信用卡欺诈检测
- Email spam filters 电子邮件垃圾邮件过滤器
- Movie or shopping recommendation systems 电影或购物推荐系统
- Facial recognition 面部识别
- Search engines 搜索引擎
- Customer service chatbots 客户服务聊天机器人
Today, companies around the world are investing heavily in researching and implementing Narrow AI for better decision-making and problem-solving. But there are some limitations to it. It requires a large amount of high-quality data to implement Narrow AI successfully. Also having quality data doesn’t guarantee perfect results. Companies have to gather, test, train, and deploy their AI models before announcing new inventions. As ANI systems are limited to performing single tasks, these systems perform inefficiently if the task changes. But replacing humans with machines can sometimes lead to frustration and lower customer experience.
如今,世界各地的公司都在大量投资于研究和实施Narrow AI,以更好地决策和解决问题。 但是有一些限制。 要成功实施Narrow AI,需要大量高质量的数据。 同样,拥有质量数据也不能保证获得完美的结果。 在宣布新发明之前,公司必须收集,测试,训练和部署其AI模型。 由于ANI系统仅限于执行单个任务,因此,如果任务发生更改,这些系统将无法有效执行。 但是用机器代替人有时会导致沮丧和降低客户体验。
Every application of machine intelligence that we’ve today is built on Narrow AI. We should not assume that these are called “Weak” because of their limited ability to perform human-like tasks. But these are called “Weak” because they’re not self-aware, self-conscious and intelligent, and driven by emotions as we humans are. Their ability to perform tasks is very limited. Even after feeding a vast amount of data, they’re nowhere close to achieving human intelligence. Or simply put, they’ll never be able to achieve intelligence like humans.
我们今天所拥有的每种机器智能应用程序都基于Narrow AI。 我们不应该将它们称为“弱”,因为它们执行类似人的任务的能力有限。 但是它们之所以被称为“弱者”,是因为它们不像我们人类那样具有自我意识,自我意识和智慧,并受情感驱动。 他们执行任务的能力非常有限。 即使在提供了大量数据之后,它们仍离实现人类智能还很近。 简而言之,他们将永远无法像人类一样获得智力。
Apple’s Siri is one of the most widely used forms of Artificial Narrow Intelligence. Siri uses natural language processing for giving recommendations or answering questions using the internet. Even though Siri can handle daily tasks efficiently, it is by no means an example of “Strong” AI. Because neither Siri has self-awareness nor possesses human intelligence. Although Narrow AI sometimes outperform humans in performing specific tasks, they fail as soon as they’re given a situation outside of their range. Even something as complex as self-driving cars is considered as Weak AI, but then the thing is that they are made up of multiple ANI systems.
苹果公司的Siri是最广泛使用的人工智能形式之一。 Siri使用自然语言处理来通过互联网提供建议或回答问题。 尽管Siri可以有效地处理日常任务,但这绝不是“强大” AI的示例。 因为Siri既没有自我意识也没有人类智慧。 尽管Narrow AI在执行特定任务时有时会胜过人类,但一旦遇到超出其范围的情况,它们就会失败。 甚至像自动驾驶汽车这样复杂的东西也被认为是弱AI,但事实是它们由多个ANI系统组成。
![Image for post](https://miro.medium.com/max/9999/1*BxR6V6pXRpJteaQirloVvw.jpeg)
Benefits, Advantages, & disadvantages of ANI –
ANI的优点,优点和缺点–
Although the ANI systems today are considered as Weak AI, they are not weak by any means. We’ve been able to successfully implement these systems in our life after years of research and efforts.
尽管当今的ANI系统被认为是弱AI,但它们绝不弱。 经过多年的研究和努力,我们已经能够成功地在生活中实施这些系统。
The important thing to remember here is that these ANI systems only perform when they’re provided with a large amount of data. They discover hidden patterns in data and extract meaningful insights from it to provide a competitive edge over others. They improve the overall efficiency, quality, and productivity of an organization and ultimately of an individual who performs all these tasks. Systems like IBM’s Watson are now able to harness the power of data combining with the latest tech of AI to assist people in customer service, finance, healthcare, and much more for better decision-making and problem-solving. Narrow AI has made our boring lives easier. That is why we shouldn’t consider it as a Weak and underestimate what it can do.
这里要记住的重要一点是,这些ANI系统仅在提供大量数据时才执行。 他们发现数据中的隐藏模式,并从中提取有意义的见解,以提供与他人相比的竞争优势。 它们提高了组织以及最终执行所有这些任务的个人的整体效率,质量和生产率。 IBM Watson之类的系统现在能够利用数据的力量与最新的AI技术相结合,来协助客户服务,财务,医疗保健等领域的人们更好地制定决策和解决问题。 狭窄的AI使我们无聊的生活变得更加轻松。 这就是为什么我们不应将其视为弱项,而低估了它的作用。
2.人工智能或强人工智能 (2. Artificial General Intelligence or Strong AI–)
Artificial General Intelligence, or AGI, or Strong AI, is the next step in the evolution of AI. It is the hypothetical ability of the machines to understand and perform any intellectual task that human beings can. It possesses the ability to exhibit human intelligence. Today’s AI is speculated to be years away from AGI.
人工智能(AGI)或强大AI是AI发展的下一步。 这是机器理解并执行人类可以执行的任何智力任务的假设能力。 它具有展示人类智力的能力。 据推测,当今的AI距AGI尚有数年之遥。
The primary goal of the AGI is to develop AI to such an extent that machines’ intellectual capabilities are equal to humans. As of now, AGI is theoretical. There are no known examples of AGI. It is believed that we will be able to achieve AGI soon. There are multiple tests to confirm human-level AGI –
AGI的主要目标是将AI开发到一定程度,以使机器的智力能力与人类相同。 到目前为止,AGI是理论上的。 没有已知的AGI示例。 相信我们很快就能实现AGI。 有多种测试可以确认人类级别的AGI –
1.图灵测试 (1. The Turing Test)
Turing test, originally called an imitation game, given by Alan Turing in 1950, is a test of a machine’s ability to exhibit intelligent behavior equivalent to, or indistinguishable from, that of a human. Turing proposed that a human evaluator would judge natural language conversations between a human and a machine designed to generate human-like responses.
图灵测试最初由艾伦·图灵(Alan Turing)在1950年提出,是一种模仿游戏,它是对一台机器表现出与人类等效或无法区分的智能行为的能力的测试。 图灵建议由人类评估者来判断人类与旨在产生类似人类React的机器之间的自然语言对话。
2. The Coffee Test
2.咖啡测试
3. The Robot College Student Test
3.机器人大学生测试
4. The Employment Test
4.就业测试
As this article says –
如本文所述 –
“Strong AI’s ultimate goal is to make an intelligent computer that can think and understand, but those terms remain ambiguous and undefinable; hence, there is no general measure of “success” in the field of Strong AI. The concept of intelligence varies from person to person, and as a result, a standardized test like the Turing Test becomes less important.”
“强大的AI的最终目标是制造一款能够思考和理解的智能计算机,但是这些术语仍然模棱两可且难以确定; 因此,在强大的AI领域,没有“成功”的一般度量。 智力的概念因人而异,因此,像图灵测验这样的标准化测验变得不那么重要了。”
Artificial General Intelligence is expected to solve problems like humans. But it is difficult to solve today’s problems with the help of current technology. It needs human intelligence and interference. World-renowned scientist, Stephan Hawking has already warned us that “The development of full artificial intelligence could spell the end of the human race.”
人工智能有望解决诸如人类之类的问题。 但是,利用现有技术很难解决当今的问题。 它需要人类的智力和干扰。 世界著名科学家斯蒂芬·霍金 ( Stephan Hawking )已经警告我们: “全面人工智能的发展可能标志着人类的终结。”
![Image for post](https://miro.medium.com/max/9999/1*OrW8z8as9ofEBOdsKFBflg.jpeg)
3.人工智能或超级人工智能 (3. Artificial Super Intelligence or Super AI)
University of Oxford philosopher Nick Bostrom defines Artificial Super Intelligence as –
牛津大学哲学家尼克·波斯特伦 ( Nick Bostrom)将人工智能定义为–
“any intellect that greatly exceeds the cognitive performance of humans in virtually all domains of interest”.
“在几乎所有感兴趣的领域中任何远远超过人类认知能力的智力”。
Artificial Super Intelligence, or Super AI, or Superintelligence, is the ability of the machine intelligence to surpass human intelligence. It is still unpredictable whether we will have human-level intelligence in the coming months, years, decades, or later this century, or until the world lasts. Regardless of the time, once the human-level machine intelligence is achieved, that super intelligent system will greatly exceed all the cognitive performances of humans in virtually all domains of interest. Such a system would be very difficult to control.
人工超级智能或超级AI或超级智能是机器智能超越人类智能的能力。 在本世纪的未来数月,数年,数十年或以后,或者直到世界持续存在之前,我们是否将拥有人类水平的情报仍是不可预测的。 无论何时,只要达到了人类级别的机器智能,该超级智能系统就将大大超越人类在几乎所有感兴趣领域中的所有认知性能。 这样的系统将很难控制。
As of now, we don’t have the Super AI and there are no examples of machines killing thousands of people. Many experts say that we don’t have to worry about ASI yet. Once it is developed, it will excel at everything — math, computers, science as well as human emotions, self-awareness, decision-making, and problem-solving capabilities. Even the brightest minds of the world can’t compete with the abilities of Super AI.
到目前为止,我们还没有Super AI,也没有任何机器可以杀死数千人的例子。 许多专家表示,我们还不必担心ASI。 一旦开发完成,它将在数学,计算机,科学以及人类情感,自我意识,决策和解决问题的能力等所有方面都表现出色。 即使是世界上最聪明的人也无法与超级AI的能力竞争。
![Image for post](https://miro.medium.com/max/9999/1*7p4n0n17CjPxPHMSl1fG7g.jpeg)
那对我们意味着什么? (So what does it mean for us?)
Artificial Intelligence research is making great progress towards its long-term goal of achieving human-like intelligence. And if it succeeds, it will be disastrous for the human race. AI’s solid growth has made many industry experts worried about AI’s take over the world. Surely it possesses the capability to become more powerful than anything the world has ever witnessed.
人工智能研究正朝着实现类人智能的长期目标迈进。 如果成功,对人类将是灾难性的。 AI的强劲增长已使许多行业专家担心AI占领了世界。 当然,它具有变得比世界上所见过的任何事物都强大的能力。
Nick Bostrom has expressed worries about robots taking over humans in his book Superintelligence. Many people like Bill Gates, Stephan Hawking, Elon Musk, Mark Zuckerberg has also raised concerns about ASI taking over the world from time to time. Even Alan Turing has expressed concerned about this –
尼克·博斯特罗姆(Nick Bostrom)在他的“ 超级 智能”一书中对机器人接管人类表示担忧。 比尔·盖茨,斯蒂芬·霍金,埃隆·马斯克,马克·扎克伯格等许多人也提出了对ASI不时接管世界的担忧。 即使阿兰·图灵曾表示关心这个 -
“If a machine can think, it might think more intelligently than we do, and then where should we be? Even if we could keep the machines in a subservient position, for instance by turning off the power at strategic moments, we should, as a species, feel greatly humbled…. This new danger…is certainly something which can give us anxiety.”
“如果一台机器可以思考,它可能会比我们更聪明地思考,那么我们应该在哪里? 即使我们能够将机器保持在从属位置,例如通过在关键时刻关闭电源,作为一个物种,我们也应该感到非常谦卑……。 这种新的危险……肯定会使我们感到焦虑。”
French Computer Scientist Yann LeCun, a pioneer of deep learning and director of AI research at Facebook says –
法国计算机科学家Yann LeCun ,深度学习的先驱和Facebook AI研究总监说 –
There is no reason for AIs to have self-preservation instincts, jealousy, etc…. AIs will not have these destructive “emotions” unless we build these emotions into them.
认可机构没有理由具有自我保护的本能,嫉妒等……。 除非我们将这些情感融入其中,否则AI不会具有这些破坏性的“情感”。
It further says,
它进一步说,
Unfortunately, it doesn’t matter whether we build in “emotions” or “desires” such as self-preservation, resource acquisition, knowledge discovery, or, in the extreme case, taking over the world. The machine is going to have those emotions anyway, as subgoals of any objective we do build in — and regardless of its gender.
不幸的是,无论我们建立在“情感”还是“欲望”中,例如自我保护,资源获取,知识发现,或者在极端情况下占领世界,都没有关系。 无论如何,机器都会产生这些情感,就像我们建立的任何目标的子目标一样,而不论其性别如何。
Although we’re still decades away from Superintelligence, AI researchers predict that the only way to achieve ASI is through AGI. No one knows when the first human-intelligence enabled machine will be developed. As Narrow AI is getting astonishingly capable, we can think of a future ruled by both machines and humans and one in which we humans are much more intelligent than machines.
尽管距离超级智能还差几十年,但AI研究人员预测,实现ASI的唯一方法是通过AGI。 没有人知道何时将开发第一台具有人类智能的机器。 随着Narrow AI的功能日渐强大,我们可以想到机器和人类共同统治的未来,人类比机器更加智能。
Thanks for reading!
谢谢阅读!
ai 项目经理 要了解