The Consciousness Conundrum: Can We Train AI Without Training Ourselves? Microsoft's AI chief Mustafa Suleyman has written an essay criticizing Anthropic's approach to training its chatbot Claude.
This method involves simulating human like consciousness in the AI, which Suleyman argues could have disastrous implications for humanity.
At the heart of Suleyman's critique is the concern that making an AI seem conscious risks creating a synthetic species with unprecedented intelligence and capability.