A4讲座预告｜黑箱内外——当代神经网络的艺术探索

莱昂-艾蒂安·库尔

追逐童年时期对计算机的迷恋，莱昂-艾蒂安·库尔于2015开始在柏林工业大学（Technical University of Berlin）学习数字媒体和技术。随着学习的进展，他的重心逐渐从学术方向的计算机科学研究，转移到计算机编程的创造性潜力上。在计算机图形学等创意类课程上取得成功后，他于2018年转入包豪斯大学(Bauhaus University)攻读媒体艺术计算机科学双学位。在新媒体艺术研究中的成功，使他获得了德国学术奖学金基金会的奖学金，并开始专注于人工智能方面的工作，由此顺利获得了关于人工智能反馈回路及其社会影响的本科学位。自5月起，他成为奥地利克拉根福的驻留艺术家，进一步开展机对机交流美学的艺术研究。自10月起，他将在科隆媒体艺术学院(Academy of Media Arts in Cologne)攻读研究生课程。他的作品用一种以计算机为中心的方法，探索人与机器/机器与机器的交互，寻找一种可以用来思考我们在数字化世界中的位置的新思维方式。

Leon-Etienne Kühr

Leon-Etienne Kühr began his studies at the Technical University of Berlin, where he pursued his childhood fascination for computers and started studying Digital Media and Technology in 2015. As his studies progressed his focus drifted away from the academic subjects of computer science and more towards the creative potential of computers and programming. After thriving in creative classes like computer graphics he decided in 2018 to switch to Bauhaus University to pursue a double-degree in computer science in media art. Thriving in the new found studies of media art he was offered a scholarship from the German Academic Scholarship Foundation and began to concentrate on the work with artificial intelligence passing his undergraduate degree about artificial intelligence feedback loops and their societal impact with flying colours. Since May he has been an artist in residence in Klagenfurt Austria to further conduct artistic research on the aesthetics of machine to machine communication. From October onwards he will study in the graduate program at the Academy of Media Arts in Cologne. In his works he explored human to machine / machine to machine interaction with a computer centric approach in search of new ways to think about our place in a digitised world.

乌苏拉·达姆与莱昂-艾蒂安·库尔两位艺术家将介绍他们各自针对人工智能算法这一主题所展开的的艺术项目，本次讲座将从两个角度审视人工智能：首先，艺术家们通过图像伪影（image artifacts）提供了人工智能黑箱内部的一瞥；其次，他们展示了当以不寻常的、有趣的方式链接时，这些算法黑箱将如何把它们的操作模式带入事物的世界。

这次展览中的作品《膜》是由乌苏拉·达姆创作的的一件装置，它通过修改、调整和扩展流行的AI算法：TGAN（时间生成对抗网络）算法，来操纵两台摄像机实时产生的影像。该装置邀请参观者在操作面板上修改软件参数。除了提供色彩缤纷、嘈杂的城市景观的变化/变形/图片（第一台相机）以及参观者本身的照片（第二台相机）的欣赏外，《膜》还使观众们得以体验人工智能的审美维度：我们是否能够通过我们的界面，探索在视觉层面上人工智能的风险和机会？我们是否能够从感官上理解算法，即使其设计目的是把我们周围环境的复杂性简化至算法定义的期望值？

莱昂-艾蒂安·库尔更进一步，他不通过看算法的输出来理解它们的美学，而是专注于算法的主张，并询问当软件与软件对话时它们将构建什么样的世界。《机器文化——带马桶和水槽的浴室》是一系列关于人工智能反馈回路的艺术实验，将文本到图像和图像到文本的转换链接起来，直到两种算法找到一个共同的平衡点。由此产生的平衡往往出人意料，并且出现在最不可能的地方，从而挑战了人类对事物进行拟人化的冲动，并质疑机器与机器之间交互的未来。由于许多被使用的算法都是基于互联网数据训练的，因此生成的图像和文本反映了我们在间或的无意之中，通过自己数字化的存在所创造出的文化。机器文化可能会在开源图库和个人的度假照片之间，成为我们自己的文化。

Ursula Damm and Leon-Etienne Kühr are presenting their artworks exploring AI algorithms. Artificial Intelligence is examined from two perspectives: First, they provide a glimpse inside the AI black box by means of image artifacts, and second, they show how these algorithmic black boxes, when linked in unusual, playful ways, carry their modes of operation into the world of things.

Membrane is an installation by Ursula Damm which manipulates live footage of two cameras with the popular AI algorithm, the TGAN algorithm, which are altered, adapted and expanded. The installation invites visitors to modify the parameters of the software on an operating panel. Beside just offering the enjoyment of a colorful and noisy variation/deformation/picture of cityscapes (first camera) and the picture of the visitor itself (second camera, Membrace also allows to experience the aesthetic dimension of Artificial Intelligence: are we capable of exploring the risks and chances of AI through our interface, on a visual level? Are we capable of sensually understanding algorithms, which are by purpose designed to reduce the complexity of our surroundings to an algorithmically defined desideratum? 

Leon-Etienne Kühr goes one step further. He does not look at the outputs of the algorithms to understand their aesthetics but focuses on the claims of the algorithms and asks what kind of world they will build when software speaks to software. “Machinecultures - a bathroom with a toilet and a sink” is a series of artistic experiments on AI feedback loops chaining text-to-image- and image-to-text-transformations until both algorithms find a common equilibrium. The resulting equilibrium oftentimes surprises and is found in the most unlikely of places, thus challenging the human impulse of anthropomorphisation and questioning the future of machine-to-machine interaction. As many of the used algorithms are trained on internet-data, the resulting images and texts reflect on the culture that we create sometimes unintentionally by our digital presence. The culture of the machines might become our own: Somewhere between stock images and vacation pictures.

膜，乌苏拉·达姆，2019，“人工智能的兑现：卑弃&解脱”展出现场，@麓湖·A4美术馆，成都，中国，2022

Membrane, Ursula Damn, 2019, installaion view at AI Delivered: the Abject and Redemption, @LUXELAKES·A4 Art Museum, Chengdu, China, 2022 

艺术家乌苏拉·达姆创作的《膜》是一件由实时影像输入生成的艺术装置。观众通过神经网络直接与生成的图像互动，这被称为 TGAN（时间生成对抗网络）算法。作品界面会引导观众根据自己的好奇心和偏好，来探索计算机的“想象力”。《膜》的图像来自来一台静态摄像机和两个监控摄像头，它们观察着展览的空间，两个屏幕实时显示着拍摄场景，视频数据流的每个像素都是一个行动单元。 一个像素在程序运行期间从不同的颜色碎块中学习，并最终提供出一种颜色，它是拍摄时间段内所有出现颜色的总和。这种简单的记忆方法创造了全新的东西：对某一特定地点的运动模式的记录。

Membrane is an art installation building on a series of generative video installations with real time video input. Membraneallows the viewer to interact directly with the generation of the image by a neural network, here the so-called TGAN (Temporal Generative Adversarial Nets) algorithm. An interface allows to experience the ‘imagination’ of the computer, guiding the visitor according to curiosity and personal preferences. The images of Membrane are derived from a static video camera and two monitoring cameras observing the exhibition space. Two screens show the scenes in real-time. Each pixel of a video data stream is considered as an operational unit. One pixel learns from colour fragments during the running time of the programme and delivers a colour which can be considered as the sum of all colours during the running time of the camera. This simple method of memory creates something fundamentally new: a recording of patterns of movement at a certain location.

更多详情和活动，请咨询👇

More information, 

please scan the QR code below

人工智能的兑现：卑弃&解脱

展览门票持续发售中

点击上方门票图片

跳转小程序购票

扫描下方二维码购买

“最近过得咋样？”

门票

【票务说明】

公众参观时间：

2021年11月21日起正式对公众开放

 每周二至周日 10:00 - 18:00 开放 

（17:30后停止入场）

一人一票，仅限验票当日有效