Creative AI Meetup: November
This event will host talks from artists and researchers presenting AI technologies and their creative applications. It is organised by curator Luba Elliott.
Event schedule:
18:00 Doors open
18:30 Introduction by event curator Luba Elliott
18:35 Anne-Sofie Maerten, PhD Student, KU Leuven: Can we tailor AI images to individual taste?
19:00 JT Liu, Computational Artist: Soundwalking the Latent space
19:25 Nick Oh, Researcher at LSE & socius labs: From Bits to Atoms: Rebuilding LeNet-1 as Sculpture (artefact(s): LeNet-1)
19:50 Talks finish, networking with drinks & food
20:30 Event close
This event is organised in collaboration with Mabyduck, a London-based startup building the tools for evaluating generative AI through human judgement. Their work covers the design of reliable human studies, rater quality and agreement and using human feedback as a signal for model training.
—
More on talks and speakers:
Anne-Sofie Maerten, PhD Student, KU Leuven: Can we tailor AI images to individual taste?
AI image generation has reached remarkable realism, yet outputs often feel generic and difficult to steer toward a distinct creative vision. This stems in part from models being optimized against a hypothetical average human preference, essentially erasing individual taste. This talk explores how we can move beyond one-size-fits-all models and make models tailored to individual taste instead.
Anne-Sofie Maerten studied psychology at KU Leuven. She has a strong interest in AI and uses ML to advance our understanding of human cognition. Her PhD is situated at the intersection of AI and aesthetics. She studied how machine learning models process visual art, exploring data augmentation techniques, VLMs for art history, and personalized aesthetic benchmarks. Her recent work explores how generative models can be tailored to individual taste.
JT Liu, Computational Artist: Soundwalking the Latent space
How could we interact with AI in music, outside of a big red button approach? The talk will present nn/memoire, an installation transforming an audio model’s latent space into an interactive treasure hunt of sounds. The audience wanders in the virtual environment, as their movements transform into unique soundscapes in real time. The work explores how we could create playful interactions between AI and a broader audience, as well as how we could ‘break’ conventional AI models for creative manipulation. Additionally, JT will be presenting some current work in development for AI-assisted musical jams.
JT Liu is a London based computational artist with roots in Beijing, Vancouver and Paris, They hold an MSc in Creative Computing from University of the Arts London, with previous background in music production. Their practice centers around novel musical affordances, playfulness, and immersive computational systems, featured in NIME, IRCAM and C&C.
Nick Oh, Researcher at LSE & socius labs: From Bits to Atoms: Rebuilding LeNet-1 as Sculpture (artefact(s): LeNet-1)
In his talk, Nick will present LeNet-1, an installation made with Alex Park that was shortlisted at the CVPR 2026 Art Gallery and received the TC PAMI Art Award. It rebuilds Yann LeCun’s 1989 LeNet-1, one of the first working convolutional networks, layer for layer in transparent circuit boards, LEDs and glass, so handwritten digits move through it as slow light. He will discuss translating bits into atoms, and the gap between seeing everything and understanding it
Nick (Seungheon) Oh is a researcher at LSE and socius labs. Advised by Professor Fernand Gobet, he works across cognitive psychology, philosophy and neural networks, in the space between every field that has studied “thinking” and every machine trying to “think”. His research has been presented at conferences and workshops including NeurIPS, ICML, COLM and AIES. He treats art-making as part of the same inquiry, building works that let people watch machines "think".