Seems like desperation to convince the industry that they're still relevant in AI after the fiasco that was Llama4. Scale doesn't create foundational models,

Seems like desperation to convince the industry that they're still relevant in AI after the fiasco that was Llama4. Scale doesn't create foundational models, they compile proprietary datasets that everyone has already licensed and trained on - not a lot of reoccurring value. Maybe this slows down competitors a bit but I doubt it.