WHILE CAN BE FUN FOR ANYONE

R1 undoes the o1 mythology in two or three significant approaches. First, There's The point that it exists. OpenAI doesn't have some sort of Specific sauce that could’t be replicated. 2nd, R1 — like all of DeepSeek’s models — has open up weights (the problem with stating “open up supply” is the fact we don’t have the data that went in