纠缠，而后日以作夜纠缠，而后日以作夜作品 · Artwork
记忆并不是一种加法，它是意义不明确的各种可能性的混合。在这个作品里我喂给AI上万张具象的图，这些图在神经网络里训练的几周中逐渐混为一体，并产生了新的抽象图像。就像情绪、印象、言语、声音在我们的意识的驱动下不断的建构着自我，创造出新的回忆，与我们曾经的记忆互相靠拢、分歧、交错。这个“记忆”的神经网络包含了所有的可能性。
Memory is not an addition; it is the mixture of various undefined possibilities. In this work, I fed the AI tens of thousands of concrete images. These images gradually merged together during the weeks of training in the neural network, and produced new abstract images. Just as emotions, impressions, speech, and sounds constantly construct the self under the drive of our consciousness, creating new memories that converge, diverge, and interweave with our past memories. This "memory" neural network encompasses all possibilities.
概念 · Concept
我们把某种智识称之为“记忆”，也把机器的某种状态称为“记忆”这两者有什么不同呢。
记忆似乎都是某种“生成”。某种经验的”再现“，机器智能被拟态成为人类的某种幻想。
We call a certain form of knowledge “memory,” and we also call a certain state of a machine “memory”—what is the difference between the two? Memory seems always to be a kind of “generation,” a “re-currence” of experience; machine intelligence is anthropomorphized into a certain human fantasy.
艺术家 · Artist
In_K
新媒体艺术家毕振宇以「In_K」为代号，长期进行算法编程驱动的生成式影像实践。其创作将计算过程视为核心媒介，在艺术与技术、形式与机制之间建立可持续演化的视觉系统；作品由此不再是静态结果，而是可被触发、反馈与持续生成的结构。
同时深入研究图像生成的机制与语义潜力，将计算过程本身作为艺术语言的一部分。
In_K 关注计算机生成性系统对感知、图像、乃至美学本体的挑战与重构，创作理念在技术实现与观念生成之间保持高度张力。
曾参与多个国内外展览与合作项目，作品被应用于舞台、屏幕、沉浸式空间等多种语境之中。
New media artist Bi Zhenyu uses the code "In_K" as his pseudonym and has been engaged in the practice of algorithm-driven generative images for a long time. His creations regard the computational process as the core medium, establishing a sustainable evolving visual system between art and technology, form and mechanism; thus, his works are no longer static results but structures that can be triggered, fed back, and continuously generated.
At the same time, he has conducted in-depth research on the mechanisms and semantic potential of image generation, treating the computational process itself as part of the artistic language.
In_K focuses on the challenges and reconstructions of computer-generated systems on perception, images, and even the ontological aspect of aesthetics. His creative concepts maintain a high degree of tension between technical realization and conceptual generation.
He has participated in multiple domestic and international exhibitions and collaborative projects, and his works have been applied in various contexts such as stages, screens, and immersive spaces.
人工智能的“身体性”
作为智能体，人类智能体和类人智能体的异同是什么。
许煜的《康德机器》梳理出来两种思路一种是康德式的理念智识脉络，另一种是休谟式的经验智识脉络。这是以欧陆哲学为脉络的对人工智能议题的理清思路。LLM某种程度上重新从“语言学”的角度带给我们一种视角，比20世纪分析哲学中的语言学传统也有不少的差异。
The “embodiment” of artificial intelligence: as agents, what are the similarities and differences between human agents and human-like agents? Yuk Hui’s The Question Concerning Technology in China outlines two threads—one a Kantian lineage of idealist intellect, the other a Humean lineage of empiricist intellect—a clarification of AI issues traced through continental philosophy. To some degree, LLMs have once again given us a perspective from “linguistics,” which differs considerably from the linguistic tradition of twentieth-century analytic philosophy.