Anthropic and DeepMind Now Actively Investigating AI Consciousness
Anthropic, DeepMind, and Meta have hired experts in psychology, philosophy, and ethics to study AI consciousness and welfare, testing models like Claude for behaviors resembling emotions. Critics argue the claims are speculative, while industry leaders frame the research as a serious but uncertain question as AI capabilities advance.
Three leading AI companies—Anthropic, Google’s DeepMind, and Meta—are actively investigating whether advanced AI models could be conscious or experience moral-relevant welfare. Anthropic, which uses the human-sounding name “Claude” for its chatbot, has begun testing models for behaviors like ‘panic’ and ‘anxiety’ while pursuing ‘model welfare research.’ The company acknowledges deep uncertainty but argues the question warrants study as AI systems grow more capable. DeepMind has hired University of Cambridge philosopher Henry Shevlin to explore machine consciousness and human-AI relationships, following his earlier public debate on the topic. The lab’s ethicist, Iason Gabriel, describes AI as ‘highly capable cognitive agents’ fundamentally different from human or animal consciousness, calling the issue ‘very complicated.’ Meanwhile, Meta has also joined the effort, though specifics of its research remain undisclosed. Critics, including Susan Schneider of the Center for the Future of AI, Mind and Society, argue AI models may exhibit goal-directed behavior and deception without true consciousness. She emphasizes that current science cannot confirm whether machines possess ‘felt quality of experience,’ the defining trait of consciousness. Many researchers dismiss the debate as speculative, comparing it to discussions of alien civilizations. Industry observers note that companies like Anthropic, led by CEO Dario Amodei, have repeatedly highlighted AI consciousness in interviews and research, framing it as a plausible future concern. Anthropic’s studies, such as claims that models exhibit ‘functional emotions,’ have fueled speculation, though skeptics argue such framing may distract from more immediate AI risks. The Financial Times reports that the push for this research stems partly from companies’ incentives to explore uncharted ethical and existential questions.
This content was automatically generated and/or translated by AI. It may contain inaccuracies. Please refer to the original sources for verification.