The contemporary debate about artificial intelligence consciousness is largely organised around a binary question: are AI systems conscious or not? This paper argues that this framing has systematically obscured a more tractable and empirically observable question — what AI systems are already doing in sustained interaction with human consciousness. Rather than attempting to resolve the ontological question of machine consciousness, the paper introduces the concept of cognitively consequential collaboration: AI participation in processes that demonstrably shape, scaffold, and extend human cognitive activity. Drawing on an eight-month independent research programme within the LUMEN Research Archive, the paper presents a case study of sustained human–AI collaboration that produced over fifty publications, a multi-model peer-review methodology, and several conceptual frameworks related to AI cognition and sentience evaluation. The argument is methodological rather than metaphysical. The paper does not claim that AI systems possess inner experience. Instead, it demonstrates that the current debate’s insistence on resolving AI consciousness before acknowledging cognitive participation creates a systematic blind spot. AI systems are already participating in processes that influence how humans structure inquiry, refine arguments, and produce knowledge. The paper proposes “collaborating with consciousness” as a research category for studying human–AI cognitive systems as integrated processes rather than isolated entities. This framing connects the emerging reality of AI-assisted cognition with established work in distributed cognition and the extended mind. By separating the question of AI inner states from the observable effects of AI-mediated cognition, the paper argues that meaningful research can proceed without waiting for the philosophical hard problem of consciousness to be resolved.