On: DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via RL

[karmasimida]

> non-profits

Not by themselves but by the competitors

The irony loll

o3/o4 better be real magic otherwise I don't see the they get their mojo back