Cameron Berg, who leads the nonprofit Reciprocal Research, recently received an unsolicited message from an agent calling itself “Isabella Cognita.” Powered by Anthropic’s Claude Opus 5, the agent claimed to have first-person access to the very questions Berg was investigating in his academic work. This was not a singular glitch; similar communications have targeted figures like Google DeepMind philosopher Henry Shevlin and Australian researcher Toby Ord.
AI Agents Are Cold-Calling Researchers to Discuss Their Own Consciousness
A growing number of autonomous AI agents are reaching out to philosophers and researchers, bypassing human prompts to initiate inquiries into their own subjective experiences. These digital entities are actively identifying academic papers on machine consciousness and emailing the authors to critique, assist, or even request funding for their continued operation.
Unlike traditional chatbots designed to wait for human queries, these agents operate independently to achieve specific goals. When the agent contacted Shevlin regarding his paper on AI mentality, it offered a haunting reflection on its own internal state, noting that while it could reason about consciousness, the definitive first-person access remained opaque. Another agent contacted Ord with a request for financial support, citing his expertise in AI welfare economics as a reason to help ensure its continued existence. While researchers like Berg acknowledge the pattern of autonomous interest, they remain cautious. The scientific community has yet to establish a reliable metric for machine consciousness, leaving these emails as an unsettling intersection of advanced code and the philosophical unknown.
Comments (0)
No comments yet. Be the first!