周末快乐。难道AI真的要取代人类？go已经完了。流行音乐也快了
- 主题发起人 zhangulei
- 开始时间
周末娱乐一下。看到你的爱因斯坦的头像，就估计你很难接受量子物理的宇宙关。如果意识独立于宇宙，世界应该是唯心的。有点类似禅宗的味道:菩提本无树，明镜亦非台，本来无一物，何处惹尘埃
我觉得有点像像佛家的涅槃状态。密宗就是利用冥想打坐进入无我，试图切断肉体的感官，去探索自我意识的世界。极深度冥想的境界，怎么会是死亡定义呢
rottenmelon
资深人士
- 注册
- 2016-12-06
- 消息
- 7,505
- 荣誉分数
- 1,905
- 声望点数
- 223
AI一开始如果被设计成替代人类而不是辅助人类 那真是人类的悲哀……
- 注册
- 2003-09-30
- 消息
- 79,630
- 荣誉分数
- 23,505
- 声望点数
- 1,373
自然进化
zhangulei
干部。干是一种美德。
- 注册
- 2018-01-06
- 消息
- 6,866
- 荣誉分数
- 1,325
- 声望点数
- 223
Can AI surpass human artists in drawing this matter?
niubishini111 (34) 在 science • 6个月前
First of all, let's ask a question: In the following 12 paintings, do you know that some paintings were made by AI, and some of them were painted by human artists? (The answer will be announced at the end of the article)
Last year, one of Google’s AIs produced a painting and sold it for US$8,500. Since then, the crusade against “AI’s last land of humanity” has been heard.
One of the paintings drawn by Google AI
Especially when the algorithm developed by AI Art Lab of Rutgers University in the United States successfully deceived the eyes of most viewers, 53% of the audience misunderstood the paintings generated by the AI painting algorithm as human artists.
Artists seem to have a bit of misunderstanding about the "AI painters" that have accumulated these codes and data. In their eyes, art is inspiration and intuition. How can it be composed of cold data and code?
However, in today's AI era, the definition of art is not unique.
Silicon Rabbit learned that there are a group of people who study computer science but love art. They are looking for a bridge between art and AI. They use the results of machine learning to empower artistic creation and find new ideas in the drawing board and bytes. .
Silicon Rabbit Jun brings the story of two young lady sisters today—one of them in San Francisco on the West Coast and one in New York on the East Coast. They are all "captives" of artistic creation, but they are all obsessed with code. The artwork they created may not be understandable to you, but they are inspiring artists.
Cynthia Hua: AI Lets Hand Painted Butterfly Take Off
Ten million butterflies are dancing. Is this just an animation? No, No, No.
The moving picture of "Butterflies flying like dancing" is an artwork called "Butterfly Room." In fact, it is a work of art created by the "Interactive Art" device. A drawing station, a camera, and a projection let the hand-painted butterflies fly under the projection.
Visitors first draw a butterfly and then take a picture of it. The device uses edge detection, neural networks, and other algorithms to generate a butterfly that only flutters its wings.
Subsequently, the butterfly group will appear on the projection, and all the visitors' butterflies come together to form this effect. Even if the visitors first painted the black-and-white butterfly, the device was also automatically filled with colors, and finally completed this colorful "Wan butterfly flying."
The original creator of this project is Cynthia Hua, an artiist-technologist who lives in San Francisco. She studied media theory at Yale University. She is also keen to study how algorithms affect the online media ecosystem. She also does mass media research and strategic adjustments for companies including Showtime, Hulu, BuzzFeed Motion Pictures, and Facebook Video.
The artworks she created are similar to the "Art + Machine Learning" project in the "Butterfly Room". For example, have you ever thought of using the location of each person's Instagram account to create a work of art?
This is called Particle Swarm Maps, an abstract animated map that combines digital Cartography and Participatory Mapping, and uses social media data to reproduce public spaces.
The most interesting aspect of this art is that it does not reproduce the contours of the city based on its geographical location, but instead tracks the movement of the Instagram location tag to reproduce the “map” of the city over a period of time (eg 30 minutes to an hour).
Finally, guess what the city is?
London.
This picture is an interactive art installation called MACHINE MIXOLOGY, which was exhibited at ACUD MACHT NEU in Berlin, Germany. Visitors can request a computer bartender based on a Gaussian mixture model to automatically generate a cocktail recipe and present it in a visual way.
Cynthia currently works for the Gray Area Foundation for the Arts (GAFA), a San Francisco-based non-profit technology arts incubator that develops and showcases a variety of avant-garde art projects that combine technology and art, including the 2016 show. Art generated by Google DeepDream.
Many researchers may use purely AI to imitate the style of painting, but Cynthia believes that the way artists use machine learning is not purely for efficiency.
In an interview with Venturebeat, Cynthia said that the most exciting type of art in science and technology is not just trying to copy the creative process through art. The artist tried to explore whether he could use machine learning and other more advanced methods to let people feel inspired and excited from the painting.
One of Cynthia's latest art projects is called A World We Made Up. In a painting similar to a map of the world, a variety of buildings with rich regional characteristics are piled up.
If you look closely, you will see Dubai's sailing hotels; fishing boats in the Philippines; office buildings in Nanjing; and distinctive buildings around the world.
This is a crowdsourcing project: Cynthia has created a website where anyone can draw a picture of your neighborhood. Subsequently, these buildings will be automatically stacked on a world map based on their geographic location.
This piece of art will eventually collect more than 10,000 works from users all over the world and then show it in June this year.
Sougwen Chung: Painting Duo with Robots
In the AI movie Machine Butler, there is such a scene: Robot Andrew sits next to Miss Amanda and plays piano music together.
And there is such a robot around Chinese artist Sougwen Chung. When Sougwen draws markers on white paper every time he picks up the marker, the robot will perform a "painting duo" on the same panel with her.
This machine is a painting robot arm called DOUG (Drawing Operations Unit Generation). It was created by Sougwen and a developer called Yotam Mann to study the possible forms of collaboration between artists and robotic arms. How to coordinate automation, autonomy, and collaborative behavior.
In 2015, the first generation of D.O.U.G was born. The first generation is not particularly smart. It observes and imitates the painter's movements through cameras mounted on the ceiling. It is mainly based on computer vision technology and creates a "human-computer interactive art" that is synchronized in real time.
The first generation of paintings did not look ideal, nor did they have a sense of art. It was more like a graffiti on the drawings. During the research process, Sougwen spent a lot of time correcting the manipulator's mimicry.
But at the time, this idea soon caught the attention of the industry, and D.O.U.G was also shown at many art exhibitions and southwest southwest (SXSW), the largest arts and technology event in the United States. D.O.U.G has also won many awards, including the 2016 Japan Media Arts Exhibition Excellence Award.
Two years later, D.O.U.G 2.0 came out. This time, the robot arm incorporates a neural network algorithm. This allows the robot arm not only to visually simulate, but to learn Sougwen's visual style of drawing, and to express the machine's expression during the “due to painting”.
Although the robotic arm primarily imitates Sougwen's movements and paintings, D.O.U.G 2.0 has been able to complete very artistic works with Sougwen due to the clearer edges and corners of the robotic arms and the immediate creation of Sougwen.
Sougwen is a Chinese artist who was born in China and grew up in Canada and currently lives in New York. Her artistic creation focuses on the human-computer interaction art, and her design also covers different fields such as installations, sculptures, still images, paintings, and performances.
She is a former researcher at the MIT Media Lab and is currently a founding member of a company (NEW INC.).
What is the relationship between AI and art?
Some people think that AI replicates art. Even an algorithm without emotion and cognition can find the rules behind the creation of human artists from the data, and then learn, imitate, and paint surrealistic masterpieces such as Picasso or Van Gogh.
Some people think that AI can help art creation. As long as the human artists “shine a little” on the drawing board, AI can immediately help them complete the next work, increase efficiency, and save time and effort.
But art geeks such as Cynthia and Sougwen do not think that AI is the brush of Linyi art, nor do they think that AI is a tool for improving the efficiency of painters. They found a third way from "paper and ink" and "nerve network" - a unique art that has never been created by AI + humans.
Similar situations are common in human history. Scientific and technological progress is not to duplicate the past development path of mankind, but to create a new pattern.
The same is true when AI collides with art.
Answer to P.S. Wenshou Question: Twelve artworks are all created by AI.
For artificial intelligence you may have other views, please pay attention to the message below。
niubishini111 (34) 在 science • 6个月前
First of all, let's ask a question: In the following 12 paintings, do you know that some paintings were made by AI, and some of them were painted by human artists? (The answer will be announced at the end of the article)
Last year, one of Google’s AIs produced a painting and sold it for US$8,500. Since then, the crusade against “AI’s last land of humanity” has been heard.
One of the paintings drawn by Google AI
Especially when the algorithm developed by AI Art Lab of Rutgers University in the United States successfully deceived the eyes of most viewers, 53% of the audience misunderstood the paintings generated by the AI painting algorithm as human artists.
Artists seem to have a bit of misunderstanding about the "AI painters" that have accumulated these codes and data. In their eyes, art is inspiration and intuition. How can it be composed of cold data and code?
However, in today's AI era, the definition of art is not unique.
Silicon Rabbit learned that there are a group of people who study computer science but love art. They are looking for a bridge between art and AI. They use the results of machine learning to empower artistic creation and find new ideas in the drawing board and bytes. .
Silicon Rabbit Jun brings the story of two young lady sisters today—one of them in San Francisco on the West Coast and one in New York on the East Coast. They are all "captives" of artistic creation, but they are all obsessed with code. The artwork they created may not be understandable to you, but they are inspiring artists.
Cynthia Hua: AI Lets Hand Painted Butterfly Take Off
Ten million butterflies are dancing. Is this just an animation? No, No, No.
The moving picture of "Butterflies flying like dancing" is an artwork called "Butterfly Room." In fact, it is a work of art created by the "Interactive Art" device. A drawing station, a camera, and a projection let the hand-painted butterflies fly under the projection.
Visitors first draw a butterfly and then take a picture of it. The device uses edge detection, neural networks, and other algorithms to generate a butterfly that only flutters its wings.
Subsequently, the butterfly group will appear on the projection, and all the visitors' butterflies come together to form this effect. Even if the visitors first painted the black-and-white butterfly, the device was also automatically filled with colors, and finally completed this colorful "Wan butterfly flying."
The original creator of this project is Cynthia Hua, an artiist-technologist who lives in San Francisco. She studied media theory at Yale University. She is also keen to study how algorithms affect the online media ecosystem. She also does mass media research and strategic adjustments for companies including Showtime, Hulu, BuzzFeed Motion Pictures, and Facebook Video.
The artworks she created are similar to the "Art + Machine Learning" project in the "Butterfly Room". For example, have you ever thought of using the location of each person's Instagram account to create a work of art?
This is called Particle Swarm Maps, an abstract animated map that combines digital Cartography and Participatory Mapping, and uses social media data to reproduce public spaces.
The most interesting aspect of this art is that it does not reproduce the contours of the city based on its geographical location, but instead tracks the movement of the Instagram location tag to reproduce the “map” of the city over a period of time (eg 30 minutes to an hour).
Finally, guess what the city is?
London.
This picture is an interactive art installation called MACHINE MIXOLOGY, which was exhibited at ACUD MACHT NEU in Berlin, Germany. Visitors can request a computer bartender based on a Gaussian mixture model to automatically generate a cocktail recipe and present it in a visual way.
Cynthia currently works for the Gray Area Foundation for the Arts (GAFA), a San Francisco-based non-profit technology arts incubator that develops and showcases a variety of avant-garde art projects that combine technology and art, including the 2016 show. Art generated by Google DeepDream.
Many researchers may use purely AI to imitate the style of painting, but Cynthia believes that the way artists use machine learning is not purely for efficiency.
In an interview with Venturebeat, Cynthia said that the most exciting type of art in science and technology is not just trying to copy the creative process through art. The artist tried to explore whether he could use machine learning and other more advanced methods to let people feel inspired and excited from the painting.
One of Cynthia's latest art projects is called A World We Made Up. In a painting similar to a map of the world, a variety of buildings with rich regional characteristics are piled up.
If you look closely, you will see Dubai's sailing hotels; fishing boats in the Philippines; office buildings in Nanjing; and distinctive buildings around the world.
This is a crowdsourcing project: Cynthia has created a website where anyone can draw a picture of your neighborhood. Subsequently, these buildings will be automatically stacked on a world map based on their geographic location.
This piece of art will eventually collect more than 10,000 works from users all over the world and then show it in June this year.
Sougwen Chung: Painting Duo with Robots
In the AI movie Machine Butler, there is such a scene: Robot Andrew sits next to Miss Amanda and plays piano music together.
And there is such a robot around Chinese artist Sougwen Chung. When Sougwen draws markers on white paper every time he picks up the marker, the robot will perform a "painting duo" on the same panel with her.
This machine is a painting robot arm called DOUG (Drawing Operations Unit Generation). It was created by Sougwen and a developer called Yotam Mann to study the possible forms of collaboration between artists and robotic arms. How to coordinate automation, autonomy, and collaborative behavior.
In 2015, the first generation of D.O.U.G was born. The first generation is not particularly smart. It observes and imitates the painter's movements through cameras mounted on the ceiling. It is mainly based on computer vision technology and creates a "human-computer interactive art" that is synchronized in real time.
The first generation of paintings did not look ideal, nor did they have a sense of art. It was more like a graffiti on the drawings. During the research process, Sougwen spent a lot of time correcting the manipulator's mimicry.
But at the time, this idea soon caught the attention of the industry, and D.O.U.G was also shown at many art exhibitions and southwest southwest (SXSW), the largest arts and technology event in the United States. D.O.U.G has also won many awards, including the 2016 Japan Media Arts Exhibition Excellence Award.
Two years later, D.O.U.G 2.0 came out. This time, the robot arm incorporates a neural network algorithm. This allows the robot arm not only to visually simulate, but to learn Sougwen's visual style of drawing, and to express the machine's expression during the “due to painting”.
Although the robotic arm primarily imitates Sougwen's movements and paintings, D.O.U.G 2.0 has been able to complete very artistic works with Sougwen due to the clearer edges and corners of the robotic arms and the immediate creation of Sougwen.
Sougwen is a Chinese artist who was born in China and grew up in Canada and currently lives in New York. Her artistic creation focuses on the human-computer interaction art, and her design also covers different fields such as installations, sculptures, still images, paintings, and performances.
She is a former researcher at the MIT Media Lab and is currently a founding member of a company (NEW INC.).
What is the relationship between AI and art?
Some people think that AI replicates art. Even an algorithm without emotion and cognition can find the rules behind the creation of human artists from the data, and then learn, imitate, and paint surrealistic masterpieces such as Picasso or Van Gogh.
Some people think that AI can help art creation. As long as the human artists “shine a little” on the drawing board, AI can immediately help them complete the next work, increase efficiency, and save time and effort.
But art geeks such as Cynthia and Sougwen do not think that AI is the brush of Linyi art, nor do they think that AI is a tool for improving the efficiency of painters. They found a third way from "paper and ink" and "nerve network" - a unique art that has never been created by AI + humans.
Similar situations are common in human history. Scientific and technological progress is not to duplicate the past development path of mankind, but to create a new pattern.
The same is true when AI collides with art.
Answer to P.S. Wenshou Question: Twelve artworks are all created by AI.
For artificial intelligence you may have other views, please pay attention to the message below。
- 注册
- 2003-09-30
- 消息
- 79,630
- 荣誉分数
- 23,505
- 声望点数
- 1,373
听起来你音乐很内行？大湿啊！
这么高深的问题！ 香浓说信息就是概率。接受器就是要最大化后验概率。最大化的后验概率是灵魂吗？
