According to TechCrunch AI, Delangue's central thesis hinges on a straightforward cost-benefit analysis, but this view may understate the non-economic barriers to open source adoption. In our view, the 'GitHub for AI' comparison only holds if the surrounding infrastructure—compute, tooling, and talent—becomes as commoditized as the models themselves. The looming risk isn't just a handful of companies controlling the models, but controlling the entire stack required to run them effectively. The industry should watch whether the move to open source simply reshuffles power from model providers to cloud and chip giants, rather than dispersing it.