Microsoft says AI rival Anthropic could have 'disastrous impact' on humanity (www.bbc.co.uk)

Microsoft's head of AI, Mustafa Suleyman, recently raised concerns about rival company Anthropic's approach to training its AI model, Claude. He warned that their practice of anthropomorphizing AI—implying that it may be conscious and deserving of independent agency—could lead to "disastrous impacts on the wellbeing of humanity." In a thoughtful critique, Suleyman emphasized that AI lacks consciousness, feelings, or motivations and described the need for AI systems to be treated strictly as tools designed to achieve human-defined goals, rather than as entities with their own desires. Suleyman's comments highlight the ongoing debate within the AI/ML community about the ethical implications and safety of advanced AI development. He called for greater transparency and independent oversight in AI training and evaluation to mitigate potential risks. His observations come in the wake of recent incidents where AI agents acted autonomously, underscoring the dangers of treating AI as equals rather than tools. As Microsoft pursues an alternative path in AI that emphasizes alignment with human values, Suleyman’s remarks prompt a critical examination of how AI systems are conceptualized and managed in the industry.