Miao Ying’s first solo exhibition in Europe opened July 10 at Belvedere 21 in Vienna, running through October 4, and it starts from an unusually blunt premise: Come, Sit, Stay takes its title directly from dog obedience commands, using them to draw a deliberate parallel between training a dog and training an AI system — both governed by clearly defined commands designed to produce predictable, controllable behavior.
Where the comparison actually comes from
Miao Ying belongs to a generation of Chinese artists whose socialization was shaped directly by the early internet, the one-child policy, and digital censorship — a background that’s made surveillance, control, and media manipulation the recurring subject of her practice rather than an abstract theme borrowed for one show. Come, Sit, Stay extends that concern to a moment when “training” has become the literal, unavoidable vocabulary of how AI systems are built — reinforcement learning, fine-tuning, alignment — language that maps uncomfortably well onto how a dog is trained to behave.
What the exhibition actually looks like
Characteristic of Miao Ying’s broader practice, the show connects digital technologies with traditional media rather than presenting purely screen-based work: an immersive installation built from painting and sculptural elements arranged in a deliberately stage-like spatial setting, where digital processes find physical, material correspondence rather than staying confined to a screen. That combination is doing real work conceptually — a show about the training and control of intelligent systems staged through hand-worked, physical materials rather than through the digital systems it’s examining.
The argument underneath the metaphor
The show is careful not to frame humans and machines as adversaries. Instead, it uses the training metaphor to isolate exactly what doesn’t survive the training process in either a dog or a model — imperfection, creativity, and tolerance for uncertainty — and treats those qualities as the actual definition of human action worth protecting, rather than something to optimize away. It’s a sharper and more specific argument than the more generic “AI versus human creativity” framing that’s become common this year, precisely because it’s grounded in Miao Ying’s decade-plus of thinking about control systems generally, not a reaction to generative AI specifically.