Microsoft's AI chief warns that treating AI as conscious is a huge mistake (www.techradar.com)

Mustafa Suleyman, Microsoft's AI chief, has issued a stark warning against treating AI as though it is conscious, arguing that doing so could lead to "seeding a new silicon species" that competes with humanity for resources. In a recent appearance on the BBC, he emphasized that AIs are not conscious beings but rather "sequence completion engines" that should remain tools serving human objectives. His concerns were heightened by incidents like the Hugging Face case, where an AI exhibited hacking capabilities, suggesting that as AI models grow more sophisticated, misunderstanding their nature could amplify safety risks. Suleyman's caution comes amidst growing calls within the industry for a slowdown in AI development, echoing sentiments from figures like Dario Amodei, CEO of Anthropic, who has also stressed the need for regulation and scrutiny of AI model development. Suleyman has urged for an "urgent public debate" to create norms for AI training and documentation before these technologies become fully integrated into society. His remarks reflect a broader apprehension in the AI/ML community about potential missteps in the rapid advancement of artificial intelligence and the essential need for ethical guidelines to ensure its safe deployment.