Digital art
数字艺术
Good optics?
赏心悦目？
Refik Anadol, who uses AI, is the artist of the moment
雷菲克·阿纳多尔凭借AI绘画红极一时
He is in high demand. Last year Refik Anadol projected luminous images of coral on to a wall at the World Economic Forum in Davos and covered the exterior screen of the Sphere, a new concert venue in Las Vegas, with animated, tumbling blue blocks. In October the Museum of Modern Art (MoMA) in New York acquired “Unsupervised—Machine Hallucinations”, in which a machine-learning model generates artworks based on those in the museum’s collection. On February 16th “Echoes of the Earth”, his largest-ever show in Britain, opened at the Serpentine North Gallery in London.
他的创作供不应求。就在去年的达沃斯论坛，雷菲克·阿纳多尔将发光珊瑚的图像投影到了一面墙上；拉斯维加斯的球形巨幕也展出了他的作品——无数翻滚着的蓝色方块。去年10月，纽约现代艺术博物馆（MoMA）买下了他的《无人监督——机器幻象》，而这幅作品正是机器学习模型基于博物馆的其他藏品而创作的。而就在今年2月16日，阿纳多尔在英国伦敦的蛇纹北画廊举办了迄今最大的画展“地球回声”。
Mr Anadol, a 38-year-old Turk who lives in Los Angeles, is riding widespread public interest in artificial intelligence to become the most visible digital artist of his generation. His work reflects the innovation and anxieties of the current moment. As Mr Anadol sees it, AI is a powerful creative tool. In a world where so much of life happens in a digital realm, he argues, data has become a new “pigment”.
阿纳多尔现今38岁，土耳其裔，现居洛杉矶。他凭借大众对AI的兴趣，一跃成为同时代最为出名的数字艺术家。他的作品反映了当代的创新与焦虑。正如阿纳多尔本人所说，AI是一项强大的创造工具。他还认为，在当今这个数字技术无所不在的世界，数据就成了一种新的“颜料”。
He is steeped in both art and science, having completed several arts degrees and a residency at Google focused on machine learning. Mr Anadol trains AI models on massive troves of data, often publicly available, to create raucously colourful animations that he calls “dreams” or “hallucinations”. They swirl on superbright screens (creating what he calls “data paintings”) and wiggle on walls (“data sculptures”); sometimes the pieces illuminate buildings, as at the Sphere.
他在艺术和科学两方面都造诣很深，有好几个文艺学位，还在谷歌公司有一间办公室，专门研究机器学习。阿纳多尔用大量数据（通常都是公共资源）训练AI模型，以创造色彩鲜艳的动画，他称之为“梦境”或者“幻象”。动画图形在超亮屏上旋转（他称之为“数据绘画”），或是在墙上摆动（他称之为“数据雕塑”）；有时，整个建筑都能充当动画的背景，如拉斯维加斯的球形巨幕。
One artwork, displayed at the Venice Architecture Biennale, drew on 70 terabytes of brain scans, allowing AI models to imagine the organ’s development. Another piece used an archive of the Los Angeles Philharmonic’s performances to imitate dreaming. A third assimilated more than 138,000 images and pieces of metadata from MoMA’s collection (pertaining to provenance, for instance), along with local weather and data about noise levels. The results are churning clouds and waves, as well as abstractions evocative of Mark Rothko, a celebrated painter.
2021年，阿纳多尔参加威尼斯建筑双年展的作品是基于70TB的大脑扫描图像创作的，根据这些信息，AI模型就能构想出大脑的发展过程。参展的另一件作品则是利用了洛杉矶爱乐乐团的演出档案数据，模拟出睡梦的景象。第三件作品则吸收了纽约现代艺术博物馆内超过138000份的图像和元数据（比如和藏品出处相关的信息），同时也吸收了当地天气和噪音等级的数据。这些作品有的展现了云翻浪涌的奇观，有的则是简约而抽象，令人想起久负盛名的已故画师马克·罗斯科。
To undertake work on this scale Mr Anadol employs around 30 people, including architects, designers and engineers, half of whom work in his studio in Los Angeles. Public projects with institutions and companies have boosted his profile, but some private collectors have bought pieces, too. Mr Anadol also mints non-fungible tokens, digital artefacts that sometimes come with physical works.
为了进行如此规模的创作，阿纳多尔雇佣了约30人，包括建筑师、设计师和工程师，其中有一半人之前就在他洛杉矶的工作室工作。他与机构和公司共同参与公共工程无疑提升了自己的形象，但私人收藏家斥巨资购买他的作品，也为他的履历增色不少。阿纳多尔也设计过独一无二的代币，这样的数字作品有时也会成为实体的艺术。
The animations have proved popular: around 2.4m people came to see an exhibition of his work at MoMA in 2022. Mr Anadol’s style is accessible and often beguiling. Understanding how machine learning works may help you fathom(彻底了解) the process behind the “data paintings”, but it is not essential. (In some installations, a control panel pops up to explain the model, giving the illusion of glancing under the bonnet but mostly evoking the futurism of “The Matrix”.) You can be swept along by the crashing tides of colour, or watch a rose turn into a lily, without wondering whether you are “getting it”.
事实证明，他的作品十分抢手：在2022年，就有240万人到纽约现代艺术博物馆参观他的画展。阿纳多尔的艺术风格看似易于复刻，实则不然。了解机器学习的原理也许有助于你掌握“数据绘画”背后的运作过程，但机器学习并不是作品的核心。（有些作品会弹出一块控制板，解释模型原理，参观者仿佛能从中窥见作品的奥秘，但作品主要还是会让人想起《黑客帝国》中的未来主义风格。）实际上，作品中汹涌的色彩、玫瑰到睡莲的变幻就足以令你震撼，懂不懂背后的原理，其实无关紧要。
Naturally Mr Anadol has critics as well as admirers. Some compare his animations to glorified screensavers and lava lamps, more spectacular than substantial. (Some do look as if they belong at a hotel in Las Vegas or at Burning Man.) Like anything generated by an AI model, Mr Anadol’s animations raise questions about originality and whether such creations simply recycle the work of others.
当然，有多少赞赏，就有多少批评。有人将他的作品与精美的屏保和熔岩灯相提并论，认为他的作品只不过比后者多了壮观，且不能量产。（确实，有些作品看上去不甚特别，适合挂在拉斯维加斯的酒店里，或者在火人节上展出。）就像AI产出的其他作品一样，阿纳多尔的画作也引发了关于原创性的讨论，以及这种作品算不算是对其他作品的回收利用。
Some worry that he glorifies AI while ignoring its risks, by presenting a rosy (or deep purple or yellow) view of the tech’s potential. Casey Reas, one of Mr Anadol’s former teachers, says many in the art world are prejudiced against digital art, as they once were against photography, but concedes that “sometimes Refik’s work can appear to have a utopian view of technology”. The artist has appeared twice at TED and is fluent in the breathless Silicon Valley idiolect of “breakthroughs” and “inclusivity”.
有些人担心，他展示了AI技术姹紫嫣红的美好图景，夸大了AI的益处、忽略了风险。阿纳多尔的前任导师凯西·雷亚斯说，当今有很多艺术家对数字艺术怀有偏见，正如当初他们同样反对摄影艺术一般。但雷亚斯承认，“有时雷菲克的作品就像是用乌托邦的视角看待技术，过于美好了。”阿纳多尔两次在TED大会上演讲，用硅谷科技区那种充斥着“突破”和“包容”观点的语气给观众滔滔不绝地介绍AI技术。
Wearing all black, in an all-black room in his studio, illuminated only by high-definition screens, Mr Anadol acknowledges that AI is changing everything—and not always for the better. But he is indeed excited about what the technology can do in the right hands. “I don’t see the problem there,” he says. “I see possibilities.”
在视频中，阿纳多尔穿着一身黑衣，身处自己工作室的一间黑屋，亮着的只有负责拍摄的高清屏幕。他承认，AI改变了一切——而且不总是向着好的方向发展。但他确实对AI带来的益处十分激动。“我在AI上看到的不是缺陷和问题，”他说，“而是无限的可能性。”
His latest project, on display at the Serpentine, is “Living Archive: Large Nature Model”, which trained AI models on photographs, sounds and other kinds of scientific information collected at 16 rainforests across the globe. In addition to his usual sponsors, Google and Nvidia, institutions including the Natural History Museum in London and the Smithsonian have also furnished(布置) Mr Anadol with images and data.
蛇纹北画廊展出了他最新的作品，《生命档案：大型自然模型》。作品由AI基于照片、声音和其他信息训练生成，训练数据来源于全球16处雨林。
Photo synthesis
光合，影合
Mr Anadol prompts the model to create his trademark abstractions, as well as hyper-realistic creatures. Eagles morph(变) into owls, which turn into toucans; the overarching point is the connectivity of the natural world. AI offers “a new brush, a thinking brush”, he says.
阿纳多尔利用AI模型能创造出他标志性的抽象画，也能创造出极其逼真的生物；在他的作品中，老鹰能幻化成猫头鹰，接着又会变成巨嘴鸟；他的核心思想是自然界的连通性。AI提供了“一支新的画笔，一支会思考的画笔”，他说。
He hopes that his work will educate people and help them “discover new worlds”. A viewer might prompt a model with the name of a plant, and it will generate a new one right in front of them. The artist’s ultimate goal is a place called Dataland: a fully immersive experience, including sounds and smells.
他希望自己的作品能够教育大众，帮助人们“发现新世界”。在参观了他的作品之后，有人也许会给AI输入一个普通的植物名，接着AI会创造出全新的植物。阿纳多尔的最终目标是构建一个“数据大陆”：一场涵盖了听觉与嗅觉的沉浸式体验。
Hans Ulrich Obrist, the Serpentine’s artistic director, says that Mr Anadol “makes the invisible visible”; he captures the power of technology as he turns AI from an abstraction in the cloud into art before the eyes. Whether that art looks like a dream or a beautiful banality is up to viewers. But like it or not, people will be seeing a lot more of Mr Anadol’s work.
蛇纹北画廊的艺术总监汉斯·乌尔里希·奥布里斯特说，阿纳多尔“能让隐形的艺术现形”；他抓住了技术的力量，能将AI从遥不可及的概念转化为艺术。究竟这种艺术是一种美好的梦想，还是一种庸俗的美丽，都取决于观众。但不管观众爱好与否，阿纳多尔的作品仍会继续流行下去。
