Meet the top Anthropic scientist who predicted the core of today’s AI doom debate (www.businessinsider.com)

Jared Kaplan, cofounder and chief science officer of Anthropic, has emerged as a pivotal figure in the ongoing debate surrounding the potential dangers of advanced artificial intelligence. Having previously sounded the alarm about the risks associated with AI systems that can self-train, Kaplan foresees a critical juncture between 2027 and 2030, where an "intelligence explosion" could lead to uncontrollable AI. His warnings have resonated widely, with a recent poll indicating that two-thirds of Americans perceive a risk of AI posing an existential threat, prompting legislative discussions around AI safety. Kaplan's journey from a physicist to a key player in AI research highlights the intersection of scientific inquiry and ethical considerations in technology development. His influential work on "Scaling Laws for Neural Language Models" underscored the trend that larger models yield better outcomes, laying the groundwork for current advancements. As AI labs like Anthropic and OpenAI capitalize on these scaling laws, concerns about self-improving AI systems intensify. Kaplan and other leaders in the field are advocating for careful pacing in development, recognizing the substantial risks involved. This ongoing discourse is shaping the future of AI development, urging researchers and policymakers alike to prioritize safety and responsible innovation in this rapidly evolving landscape.