Ray Tijssen Debuts AI Algorithmic Organisms in Kuala Lumpur

Most public generative-AI art installations run for a single event or a weekend; *Algorithmic Organisms 2.0* is committed to producing continuously novel audiovisual output for 16 straight days in a public venue. That duration requirement, not the visuals themselves, is the interesting engineering constraint here: it rules out a curated loop and forces a generative system that can run unattended without repeating itself or degrading.

According to Variety and event organizer Destina, Algorithmic Organisms 2.0 opens at The Grey Box, GMBB in Kuala Lumpur and runs July 4 to 19, 2026. The piece is created by Dutch audiovisual artist Ray Tijssen, who works under the name 0010x0010 and serves as Creative Director at SQNXR. The Kuala Lumpur presentation is organized by Jazzy Group of Companies in collaboration with ticketing platform MAISEAT, and is supported by the National Art Gallery of Malaysia (Balai Seni Negara). Production credits include Alien Illuminate and SQNXR, per Destina, and an invited media/partner preview was held on July 3. Eksentrika reports this is not the work's debut: an earlier version of *Algorithmic Organisms* first showed at Bangkok's Museum of Contemporary Art (MOCA) in 2023, with minimal wall text by design so visitors interpret the shifting forms themselves.

Public reporting describes the installation's behavior but not its underlying stack: continuously evolving visuals, sound, and digital forms with no fixed start or end point, generated live rather than played back from a fixed loop. No outlet names the specific models, engines, or render hardware in use. Industry-pattern observation: comparable multi-week generative installations typically pair locally hosted, low-latency rendering (to avoid network jitter breaking the audiovisual sync) with a constrained generative model tuned for bounded novelty, since a fully unconstrained model risks producing jarring or off-brand output during unattended operation.

The operationally interesting problem is reliability over duration, not the generation technique itself. Running a generative pipeline continuously for 16 days in a public gallery, with no operator actively curating each frame, requires graceful degradation when a model or renderer produces a bad output, session-level state management so the piece feels continuous rather than looping, and monitoring/alerting equivalent to a production service rather than a one-time demo. Teams building long-running generative or agentic experiences for public or customer-facing deployment face the same class of problem, just with different stakes.

None of the current sources disclose the model or runtime architecture. Artist talks, MAISEAT program notes, or a post-show technical writeup could surface details on model choice, on-prem vs. cloud inference split, and how failure states are handled during the 16-day run - any of which would be more valuable to practitioners than the current experiential coverage.