anthropic agility claude code stickiness chinese tools

Anthropic has no real advantage without AGI. The whole thing is just betting on AGI. Claude Code isn't very sticky; it's easy to replace with tools like Codex, Cursor, or Devin at much cheaper prices. It lacks the stickiness of traditional SaaS because it doesn't have data gravity. Opus is good, but other coding tools are always catching up. Chinese coding tools are pretty close in performance but much cheaper since Anthropic faces high costs for electricity and data centers. The Chinese face similar issues but can keep cutting prices because their electricity costs are much lower. Their data centers have to be in the US due to data concerns. The only edge they have is that big western companies avoid Chinese models, though smaller companies might still use them. Their only hope is AGI. Without it, their revenue will stall, and they'll keep spending on electricity, AI research, and chips to train new models, leading to poor margins even if costs drop over time. They know this, so they're trying to expand into new areas like design and finance, but I don't see much progress. If they build a full suite of Claude tools for every business area and sell it as a package, that might work. But since AGI isn't achievable just by scaling LLMs, that seems unlikely.