I want you to envision a single piece of artwork generated by artificial intelligence. When most of us think of AI art, I bet we're imagining something like this. We're all probably picturing something totally different.
我想请你想象一件 由人工智能生成的艺术品。 大多数人想到 AI 艺术的时候, 我相信我们想到的是这样的东西。 我们可能都在想 一些完全不同的东西。
Today, with machine learning models like DALL-E, Stable Diffusion and Midjourney, we've seen AI produce everything from strange life forms to imaginary influencers to entirely foreign, curious kinds of imagery. AI as a technology is fascinating to us because we're inherently drawn to things we cannot understand. And with neural networks processing data from thousands of other images made by people from every possible generation, every art movement, millions of images in one simple scan, they can produce visuals that are so familiar yet strikingly unfamiliar. More poetically, AI mirrors us.
如今,借助 DALL-E、Stable Diffusion 和 Midjourney 等机器学习模型, 我们已经看到 AI 生成了 奇怪的生命体、 虚拟网红 和各种陌生的、奇怪的图像, AI 这项技术令我们着迷, 因为我们天生 就会被无法理解的事物所吸引。 神经网络可以处理上万张图片的数据, 来自每一代人、每一场艺术运动, 数百万张图像 仅凭“一键扫描”即可完成, 它也可以生成我们既熟悉 又非常陌生的图像。 更诗意地说, AI 是我们的一面镜子。
The world is beginning to change right before our very eyes, and it's basically divided into two schools of thought. There are pessimists who think AI poses a great threat to human creativity. And then optimists who see it as an extension of our creativity. So is it even possible to be truly original as an artist anymore? How do we begin to critically engage with artworks made by machines?
世界在我们眼前变化, 现在分裂成两个思想流派。 悲观主义者认为 AI 对人类的创造力构成了巨大威胁。 乐观主义者 视 AI 为人类创造力的延伸。 作为一名艺术家, 是否还有可能真正做到原创? 我们又该如何开始深入 接触机器制作的艺术品?
We can start by looking at some metaphors, narratives and insights from artists who are truly pushing the boundaries of AI. Let's look to these moments of delight, surprise, confusion and wonder that give us just one small glimpse into the possibilities of encounter with this technology. Because as we've seen, this is a very moral and ethical encounter as much as an aesthetic one.
我们可以从真正突破 AI 界限的艺术家 所使用的一些隐喻、 叙事和见解开始。 让我们来看看这些 充满欢乐、惊喜、 困惑和惊奇的时刻 , 它们让我们瞥见了 与这项技术接触的不同可能性。 正如我们现在所见, 这是一次道德和伦理的相遇, 也是一次美学上的遇见。
Mario Klingemann sold this piece on auction in 2019. It is running an AI model trained on thousands of portraits from the 17th to 19th centuries. The model constantly reveals uncanny interpretations of the human face. Each one is unique, generated in real time as the machine reads its own output. For the viewer, it's almost like peering into the machine's hallucinations as it conjures each new portrait.
马里奥·克林格曼(Mario Klingemann) 在 2019 年拍卖会上出售了这件作品。 这个 AI 模型的训练材料是 17 至 19 世纪的数千幅肖像画。 该模型不断呈现着 对人脸的诡异诠释。 每一个都是独一无二的, 在计算机读取自己的输出时实时生成。 对于观众来说, 每生成一张新肖像, 就像是在窥见这台机器的幻觉。
Sofia Crespo's series "Neural Zoo" uses neural network interpretations of the real world to generate unreal sea creatures and diverse biological forms. Frogs look like flowers. Translucent jellyfish have vivid internal organs. There’s no one real creature in these images, but AI allows us to envision otherworldly lifeforms in impossible detail.
索非亚·克雷斯波(Sofia Crespo) 《神经动物园》系列, 利用神经网络对现实世界的认知 生成虚构的海洋生物 和各种生物形态。 像花一般的青蛙。 半透明的水母 有着清晰可见的内脏器官。 这些图像中没有真实的生物, AI 使我们能够以不受限的细节 设想不同寻常的生命。
This abstract piece by Sara Ludy began as a digital painting. It was augmented to fit a 16-by-9 ratio, using a prompt for "torn edges" in DALL-E 2's Outpainting. Outpainting allows artists to extend their creativity beyond the frame using simple language prompts like "torn edges."
萨拉·卢迪(Sara Ludy)的这幅抽象作品 最初是一幅数字绘画。 在 DALL-E 2 扩图的 “撕裂边缘”功能之下, 它被延展到了 16 x 9 的比例。 扩图让艺术家可以通过简单的提示 (例如 “撕裂边缘”) 将创造力扩展到画面之外。
This piece by Ivona Tau might read as a photograph, but it is also the work of AI. It's the result of GAN training on thousands of images from the artist's personal photo collection. Tau curates from her own photographs, carefully choosing the inputs and outputs for the model. In many ways, AI art is a form of curation. It becomes the process of selecting from hundreds of images at a time. This video pulls from models trained on a massive data set of Tau's photos, resulting in a kind of algorithmic memory. But she also created a destructed data set for the model to symbolize forgetting or fleeting memory.
伊沃娜·塔乌(Ivona Tau)的作品 看上去像是一张照片, 但它其实是 AI 的作品。 这是 GAN (生成对抗网络)接受了 艺术家个人照片集中的数千张图像 训练后的结果。 塔乌对自己的照片进行整理, 仔细筛选了模型的输入和输出。 从某些方面来说, AI 艺术是一种策展形式。 它是一次性从数百张图像中 进行筛选的过程。 这段视频由塔乌照片构成的 大型数据集训练的模型制作, 由此产生了一种算法记忆。 但是她还为模型创建了 一个销毁数据集, 代表遗忘或短暂的记忆。
And finally, we have Claire Silver. Silver has called herself a “collaborative AI artist” in that she works intentionally with the machine to produce her art. Her process is constantly evolving as the tools evolve. She often works with inpainting techniques, masking and transforming just one small piece of an image. For this portrait, she shifted the opacity of various sections with an Apple pencil, transforming it bit by bit. She likens this technique to her version of glazing in oil painting. Silver feeds AI-generated images from one model into another, effectively creating new forms of language and understanding for the machine itself. Her work is half master painting, half digital art. Both old and new. This piece pulls inspiration from famed artists like John Singer Sargent, Evelyn De Morgan and Gustav Klimt, almost as an homage.
最后,我们来看看 克莱尔·西尔弗(Claire Silver)。 西尔弗称自己为 “AI 合作艺术家”, 因为她有意识地 使用机器创作自己的作品。 随着 AI 工具的发展, 她的流程也在不断变化。 她经常使用图像修复技术, 来掩盖和变换图像中的一小部分。 这幅肖像中,她用苹果铅笔 改变了各个部分的透明度, 一点一点转变了它。 她把这种技巧比作 在油画中的上釉。 西尔弗将人工智能生成的图像 输入到一个个模型中, 有效地为机器本身 创造了新的语言和理解形式。 她的作品一半是大师绘画, 一半是数字艺术。 新旧融合。 这件作品从约翰·辛格·萨金特、 伊芙琳·德摩根和古斯塔夫·克里姆特 等著名艺术家那里汲取了灵感, 几乎是一种致敬。
Because different AI models are trained on different sets of information, it's almost like they're all speaking different languages. AI is everywhere now. We are all now collectively co-creating with AI, whether we're aware of it or not. If we want to be a part of these worlds, we cannot design alone. If we want to be culturally literate in these new kinds of images and predictions and forms, then looking to the work of artists is a very productive place to start. We need to brace ourselves for an increasingly technological future, which is only going to multiply all the creative possibilities at our fingertips now.
由于不同的 AI 模型是 由不同的信息集训练的, 所以就像它们说着不同的语言。 如今,AI 无处不在。 无论我们是否意识到, 我们每个人都在与 AI 共同创作。 如果我们想成为这些世界的一部分, 我们不能孤军奋战地设计。 如果我们想对这些 新的图像、预测和形式 培养一定的理解能力, 这些艺术家的作品 会是个非常有效的起点。 我们要为日益技术化的未来 做好准备, 它会增加我们指尖上 所有创造性的可能。
Thank you.
谢谢。
(Applause)
(掌声)