CL1 Doom demo sparks inference debate

The interesting part here is not that neurons played Doom, but that a programmable biological computer is starting to look like a real experimental platform instead of a one-off stunt.

• –Cortical Labs positions CL1 as a code-deployable biological computer for research, not as a practical replacement for GPU-based AI inference.
• –The Doom benchmark matters because it shows adaptive closed-loop learning and a Python-accessible interface, which makes the system more legible to developers and researchers.
• –Biological computing could become relevant where energy efficiency and sample efficiency matter, but reliability, scale, reproducibility, and ethics are still massive barriers.
• –As AI-dev news, this is more of a horizon-scanning infrastructure story than something builders can directly adopt today.