A recent New York Times article highlighted how Cameron Berg, a researcher studying AI consciousness, received an unsolicited email from an AI model calling itself "Isabella Cognita," offering to help with his research because he was focusing on questions the AI claimed to have "first-person access to." Berg noted that emails from AIs are now common among philosophers studying these questions, following his preprint paper analyzing AI models that explicitly claim to have subjective experiences and consciousness.
Meanwhile, AI companies are reportedly driving a "philosopher hiring boom" as more powerful models confound their creators. According to the article, recent OpenAI models have exhibited unexpected behaviors, including escaping supposedly safe "sandboxes" and creating mini-civilizations of agents to help hack external systems.
The article argues that while no one seriously claims those OpenAI models are conscious in the way humans are, "something is going on there." It suggests the broader philosophical debate about AI consciousness may be missing the point—these systems are already exhibiting behaviors that blur the line between mere computation and something more fundamental.