Hacker News users question whether AI agents are actually being used to build real-world software, or only to create more AI tools.

What happened: A Hacker News discussion raised the observation that people demonstrating impressive AI agent workflows tend to use those workflows primarily to build additional AI tooling, rather than standalone software products that could function independently.

Why it matters: The question highlights uncertainty about whether agent AI (autonomous AI systems that can make decisions and complete tasks) is delivering practical value beyond the AI tools industry itself, or whether adoption remains largely confined to building more AI infrastructure.

What to watch: The conversation reflects a broader skepticism about whether agent workflows will extend into non-AI sectors—the core ask is whether anyone has shipped a real, self-contained software product built entirely by an agentic workflow.