AGI Is Here (breaking-changes.blog) AI

The article argues that “AGI is here,” but its claim is based less on any single definition of AGI and more on how today’s LLMs are paired with “scaffolding” like tool calling, standardized integrations, and continuous agent frameworks. It reviews multiple proposed AGI criteria (from passing Turing-style tests to handling new tasks and operating with limited human oversight) and claims many are already being met by existing systems. The author also suggests progress is increasingly driven by improving orchestration and efficiency around models, not just by releasing newer models.