Notes from ISEA2025 – Session Two

This second session at ISEA2025 unfolded like a mirror hall —
flickering between memory, mimicry, identity, and critique.
The topic was AI art.
But the deeper current was how AI remakes our relationship to memory and self —
and what we lose when simulation begins to feel like remembrance.

Synthetic Punctum – Xuanyang Huang, Xiaoyun Zhong, David Kei-Man Yip
This talk began with a box of anonymous black-and-white photographs, found in a street market.
Faces, landscapes — no names, no metadata, no origin story.
The artists asked:

What if AI could “reconstruct” the lost moments behind them?

Using surveillance footage, Stable Diffusion, and clever prompt engineering,
they generated new images that echoed the aesthetic of the found photos.
These were then printed, giving them the tactile authority of the real.

But something troubled the room:

Whose memories were these?

What is the difference between reconstruction and fabrication?

Can AI simulate the punctum — that tiny wound of intimacy Roland Barthes described —
or does it only offer the appearance of one?

These were beautiful images.
But I left wondering:

Are we making memorials — or erasing the silence that absence once protected?

How Artists Use AI as a Responsive Material – Rozental, van Dartel, de Rooij
This presentation reframed AI not as a tool, but as a responsive material —
something artists engage with like clay, light, or breath.

Rooted in Material Engagement Theory, their framework offered four key concepts:

Radical continuity: no boundary between thought and making

Creative thinging: the material participates in thinking

Enactive discovery: ideas emerge through doing

Attentive unity: staying present with the process, not mastering it

Artists using AI, they argued, do not just generate images —
they develop an intuition with the system.
They engage in co-creation, embracing serendipity, and letting the machine misunderstand beautifully.

For me, this felt like the most alive part of the session.
It reminded me that even in code and computation,

There is space for listening, surprise, and trust in emergence.

AI Art from Latin America: Horizons of Co-Existence – Wolfgang Bongers
This was not a presentation.
This was a disruption.

Bongers asked what it means to make AI art from Latin America —
not just in terms of geography, but through radically different frameworks:

Territoriality, not data

Ancestral intelligence, not abstract logic

Resistance, not simulation

He referenced artists and thinkers like Giselle Beiguelman and Daniela Catrileo,
foregrounding practices of refusal, glitch, silence, and opacity.

The flood of web addresses he shared was not overload —
it was a counter-archive.
A refusal to let AI art be flattened into slick output and universal models.

This talk left me asking:

Can we decolonize intelligence itself?
What if AI is not a tool to be used, but a field to be resisted, repurposed, re-breathed?

Exploring Postmodernism in AI Art – Hantian Xu and James She
This final talk offered a theoretical frame:
AI art as the postmodern gesture reborn —
through appropriation, deconstruction, and poetic narrative.

Duchamp, Abramović, Calle were invoked.
LoRA models, math-heavy slides, and prompt tricks filled the room.
At one point a slide vanished. So did the speaker’s thread. But he recovered —
and maybe that was the most authentic postmodern moment of all.

Still, I left wondering:

Is this really critique — or just aestheticized collapse?
Are we repeating the gestures of postmodernism without new stakes?

Closing
This session – like the first one – left me with more questions than certainty.
But that felt right.

These people are testing the edge of knowing —
asking how much memory, authorship, and affect can survive
when the machine is asked to do the remembering.

We are not beyond art.
We are inside its reformatting.

Stay with the tension.
More to come.