In such a hot situation of artificial intelligence, there have been many examples of using AI systems to write articles, write songs, and create visual art, but the question is, can you tell whether they are created by humans or created by AI systems?

在人工智能如此炙手可热的情况下，有许多使用AI系统编写文章，编写歌曲和创建视觉艺术的示例，但问题是，您能分辨出它们是由人类创造还是由AI系统创造的？

Gaetan Hadjeres and Francois Pachet from the Sony Computer Science Laboratory in Paris trained their AI system on a chorus song by Johann Sebastian Bach. Choosing these chorus songs, as the MIT Tech Review said, “because the production process is stepwise and uses algorithms.” Training this AI system uses 352 Bach songs, which are converted into different notes, a total of 2503 Works. These works made more than half of the 1,600 listeners (including professional musicians and music students) think that they were Bach’s works.

巴黎索尼计算机科学实验室的Gaetan Hadjeres和Francois Pachet用Johann Sebastian Bach的合唱歌曲训练了他们的AI系统。 正如麻省理工学院技术评论所言，选择这些合唱歌曲是“因为制作过程是逐步的并且使用算法”。 训练该AI系统使用352首Bach歌曲，这些歌曲被转换成不同的音符，共2503作品。 这些作品使1,600位听众(包括专业音乐家和音乐学生)中超过一半的人认为他们是巴赫的作品。

If it is the field of poetry, the AI ​​system is even more challenging. In the competition held at Dartmouth College, the task of the judge is to read 14 lines of poems and the prescribed sentences of escort. Some are written by humans, and some are written by AI systems. In this case, all judges can identify which are written by humans and which are written by the AI ​​system.

如果是诗歌领域，那么AI系统将更具挑战性。 在达特茅斯学院举行的比赛中，法官的任务是朗诵14行诗和规定的押送句子。 有些是由人类编写的，有些是由AI系统编写的。 在这种情况下，所有法官都可以识别哪些是由人类编写的，哪些是由AI系统编写的。

In addition, what will happen to using AI to generate novels? A machine learning enthusiast shared his novel “Harry Potter” written after training the AI ​​system with deep learning algorithms, and the effect is relatively average.

此外，使用AI创作小说会发生什么？ 一位机器学习爱好者分享了他用深度学习算法训练AI系统后写的小说《哈利·波特》，其效果相对中等。

So, will AI systems create works that can be equivalent to human creations? The answer is: You can first understand how humans think creatively through works created by humans.

那么，人工智能系统会创造出可以与人类创造等同的作品吗？ 答案是：您可以首先了解人类如何通过人类创造的作品进行创造性思考。

Although humans’ rational thinking and mathematical abilities are innate, we still need to fully realize these abilities through education. So we study the laws of nature, logical puzzles, ethical dilemmas and so on. However, even the best of us can make strange and unreasonable decisions that give us all kinds of prejudices.

尽管人类的理性思维和数学能力是与生俱来的，但我们仍然需要通过教育充分实现这些能力。 因此，我们研究自然定律，逻辑难题，道德困境等。 但是，即使是我们最好的人，也会做出奇怪和不合理的决定，这些决定会给我们带来各种偏见。

On the other hand, human emotion, intuition, and creativity only require a little formal training. Every child will laugh, cry, draw, create, question, explore, etc., let alone education. Education is used to shape these abilities, teach children to control their emotions, guide their creativity, and create with more precise and thoughtful considerations.

另一方面，人类的情感，直觉和创造力只需要一点正规的训练。 每个孩子都会笑，哭，画，创造，提问，探索等，更不用说教育了。 通过教育来塑造这些能力，教孩子们控制情绪，引导他们的创造力，并以更加精确和周到的考虑进行创造。

Can AI learn to create art?

人工智能可以学习创造艺术吗？

The computer system is completely different. Their algorithm tasks perform well, but they lack most of the natural capabilities of humans. Of course, they can complete complex mathematical calculations in milliseconds, but what if you tell a joke to the computer or play a love song? It won’t work.

计算机系统是完全不同的。 他们的算法任务执行良好，但是却缺乏人类的大多数自然能力。 当然，他们可以在毫秒内完成复杂的数学计算，但是如果您向计算机讲个笑话或播放一首情歌怎么办？ 它不会工作。

Therefore, even with the rapid development of artificial intelligence, whether AI systems can learn to create art is an open question. However, progress in areas not directly related to creativity still means that it is still possible. Algorithms have completed some of the most difficult logical puzzles designed by humans. Artificial intelligence systems are accepting the challenge of understanding language, such as torturing and killing the best human players in games, such as taking to the streets to prove how safe they are.

因此，即使随着人工智能的飞速发展，人工智能系统是否可以学习创造艺术仍是一个悬而未决的问题。 但是，与创造力没有直接关系的领域的进步仍然意味着它仍然是可能的。 算法已经完成了人类设计的一些最困难的逻辑难题。 人工智能系统正在接受理解语言的挑战，例如折磨和杀死游戏中最优秀的人类玩家，例如上街证明自己的安全性。

Many of these skills are not far from the creativity itself. Especially language, language is very follow specific rules, but also allows artistic pursuits, such as storytelling and poetry.

这些技能中有许多与创造力本身相距不远。 特别是语言，语言非常遵循特定的规则，但也允许艺术追求，例如讲故事和诗歌。

The power of human creativity

人类创造力的力量

In view of the differences in these mental processes, the human brain is logical and creative, which is fascinating. Logic runs on a set of rules and procedures in an orderly manner. On the other hand, creativity and intuition can be messy, but straightforward.

鉴于这些心理过程的差异，人脑是富有逻辑性和创造力的，令人着迷。 逻辑以有序的方式运行在一组规则和过程上。 另一方面，创造力和直觉可能很杂乱，但很简单。

First, we should admit that there are at least two forms of creativity. We can call the ideas we think of when we take a shower in the shower room, for example, as creative examples of “a-ha”. Then, as we create, stretch our minds and improvise like a pianist. In both cases, focus and cognition are put aside, but emotional expression and subconsciousness are freely controlled.

首先，我们应该承认至少有两种形式的创造力。 例如，我们可以将在淋浴房中淋浴时想到的想法称为“哈哈”的创意示例。 然后，当我们创作时，像钢琴家一样张开胸怀，即兴创作。 在这两种情况下，都将注意力和认知放在了一边，但情绪表达和潜意识却可以自由控制。

Thoughts wandering, this state of mind is to give up conscious thought and let the mind go wherever it is. This state has been linked to the creative process more than once. Ideas are pouring out of my head, stimulating something unexpected. Unlike mathematical formulas or logical structures, these processes are not easy to recreate in the form of calculations.

思想在徘徊，这种心态是放弃有意识的思想，让思想走到任何地方。 此状态已多次与创作过程相关联。 想法从我脑海涌出，激发出一些意想不到的东西。 与数学公式或逻辑结构不同，这些过程不容易以计算形式重新创建。

Long ago, formal psychological research became creativity, and many artists were caught in the power of thought wandering. When the surrealist painter Salvador Dali was sitting in a chair, holding the key in his hand, he realized that in those happy moments before going to bed, when strange thoughts appeared, the key would fall and wake him up.

很久以前，正式的心理学研究开始具有创造力，许多艺术家陷入了思想游荡的力量。 当超现实主义画家萨尔瓦多·达利(Salvador Dali)坐在椅子上，手中握着钥匙时，他意识到在睡觉前的那些快乐时光中，当奇怪的想法出现时，钥匙会掉下来并唤醒他。

However, not every showering process is considered a creative process. Some thoughts will be ridiculous when you think about it the next day, and some melody that came up in your head will be just a harsh noise. Some more subtle things are needed for creativity to reach the state of the art. The creator must understand these rules in order to destroy them effectively. They must have a feeling, a feeling or effect that suits them when creating. In this way, they are introducing new ideas into new ideas to some extent. Artists combine things in a new way and look at them from a new perspective, using materials that we are very familiar with, such as paints, phrases, melody and so on.

但是，并非每个淋浴过程都被认为是一个创造性的过程。 第二天想到这些想法时，有些想法会很荒谬，而脑海中浮现的一些旋律只会是刺耳的声音。 为了使创造力达到最先进的水平，还需要一些更细微的事情。 创建者必须了解这些规则，才能有效地销毁它们。 在创建时，它们必须具有适合他们的感觉，感觉或效果。 这样，他们在某种程度上将新思想引入新思想。 艺术家使用一种我们非常熟悉的材料(例如绘画，短语，旋律等)以一种新的方式组合事物，并以新的视角看待它们。

Jobs said before: “Creativity is just the ability to connect things.” Although he said “just”, it is actually much harder. It needs a purpose, a point, whether it is solving a problem or describing an idea. We can teach AI systems to group things together, such as images, tones, and words, but can AI be done in meaningful and clever ways?

乔布斯之前说过：“创造力就是连接事物的能力。” 尽管他说“公正”，但实际上要困难得多。 无论是解决问题还是描述想法，它都需要一个目标，一个要点。 我们可以教AI系统将诸如图像，色调和单词之类的东西组合在一起，但是AI可以有意义且聪明的方式完成吗？

An attempt at artificial intelligence creativity

人工智能创造力的尝试

Researching out how creativity works, and teaching it to intelligent machines, although extremely challenging, did not stop researchers from trying. At the beginning, I have seen examples of writing articles and songs using AI systems.

尽管极富挑战性，但研究创造力的工作原理并将其教授给智能机器，并没有阻止研究人员尝试。 一开始，我看到了使用AI系统编写文章和歌曲的示例。

Let’s talk about the visual arts. AARON was “born” in 1973 as an AI system created by artist Harold Cohen. In an interview with the BBC, Cohen said: “AARON has become self-sufficient.” But he also said that AARON is still a long way from truly human-like creations: “I don’t deny that at some point in the future, A machine can make some artistic things more complicated. These are much more complicated things than autonomous driving, but I don’t expect it to be achieved in this century.”

让我们谈谈视觉艺术。 AARON于1973年“诞生”，是由艺术家Harold Cohen创建的AI系统。 科恩在接受BBC采访时说：“ AARON已经变得自给自足了。” 但是他还说，AARON距离真正的类人作品还有很长一段路要走：“我不否认在将来的某个时候，一台机器会使某些艺术作品变得更加复杂。 这些比自动驾驶要复杂得多，但是我不希望在本世纪实现。”

Appreciate the art of machine creation

欣赏机器创作的艺术

An important question remains: How do we feel the artwork made by smart machines? Will we appreciate its creativity and design?

一个重要的问题仍然存在：我们如何看待智能机器制作的艺术品？ 我们会欣赏它的创意和设计吗？

There is a school of thought that the value and significance of art are independent of its creator. On the other hand, many people believe that the story behind the art and information about the artist may influence our perception of the work of art. You can take a test like this: Imagine you are standing in front of a very famous work of art. Then someone tells you that this work is just a fake. Even if it is exactly the same as the original, do you still find it fun to appreciate it?

有一种流派认为艺术的价值和意义独立于其创造者。 另一方面，许多人认为艺术背后的故事和有关艺术家的信息可能会影响我们对艺术品的看法。 您可以进行如下测试：假设您正站在一幅非常著名的艺术品面前。 然后有人告诉你，这项工作只是伪造的。 即使它与原始照片完全一样，欣赏它仍然很有趣吗？

Psychologist Paul Bloom pointed out in an interview that when showing people an object or face, what people value most is “people’s evaluation of it”, how do you evaluate them to describe the object or face? It has a great impact on them. A violinist played a piece of music at a subway station in Washington, and received $32 in consumption. No one told passers-by that the violinist is Joshua Bell, who has released more than 30 albums and performed in the White House. At Bell’s concerts, the average ticket price per ticket was more than $100! If passers-by knew his identity, would more people come to listen to the music he played at the subway station? The answer is very obvious.

心理学家保罗·布鲁姆(Paul Bloom)在一次采访中指出，当向人们展示物体或面Kong时，人们最看重的是“人们对物体或面Kong的评价”，您如何评价他们来描述物体或面Kong？ 这对他们有很大的影响。 一位小提琴手在华盛顿的一个地铁站演奏音乐，并获得了32美元的消费。 没有人告诉过路人，小提琴手是约书亚·贝尔，他发行了30多张专辑并在白宫演出。 在贝尔的音乐会上，每张门票的平均票价超过100美元！ 如果路人知道他的身份，会不会有更多的人来听他在地铁站播放的音乐？ 答案非常明显。

It will take a while for the AI ​​system to learn creativity

人工智能系统需要一段时间才能学会创造力

As it is now, creativity is still human behavior. Although many AI systems are trying to create works of art, they still seem to fall short of the artist’s standards. It’s a good performance in music works, but in most other fields, the algorithm is still not satisfactory. But one thing is certain: as machines become smarter and more capable, they will get closer to true creativity.

现在，创造力仍然是人类的行为。 尽管许多AI系统都在尝试创作艺术品，但它们似乎仍未达到艺术家的标准。 它在音乐作品中表现不错，但在其他大多数领域，该算法仍不令人满意。 但是有一点可以肯定：随着机器变得越来越智能，功能越来越强大，它们将越来越接近真正的创造力。