Is AI Slowing Down? Nathan Labenz Says We're Asking the Wrong Question

AI Development & Progress - The discussion revolves around whether AI progress is slowing down or if society is simply adapting to the rapid advancements [1] - The debate includes considerations around GPT-5, its potential capabilities, and how its launch perception might impact the industry [1] - Scaling laws are mentioned in the context of GPT-4.5 and the changes anticipated with GPT-5, suggesting a focus on model scaling and its effects [1] - Longer context windows and improved reasoning capabilities are highlighted as key areas of AI development [1] AI Applications & Impact - AI's potential role as a scientist and its ability to facilitate new discoveries are explored [1] - The discussion covers the future of coding, AI agents, and recursive self-improvement, indicating an interest in AI's transformative impact on software development and automation [1] - The potential of multimodal AI and robotics beyond chatbots is considered, suggesting a broader view of AI's applications [1] Societal & Economic Implications - The conversation addresses the impact of AI on jobs and automation, referencing the METR study, which is often misunderstood [1] - The importance of building a positive vision for the future of AI is emphasized, highlighting the need for ethical considerations and responsible development [1]