There are already open-weight models that match OpenAI's best models (at the time of their release), so the idea that OpenAI would lose something by making its

There are already open-weight models that match OpenAI's best models (at the time of their release), so the idea that OpenAI would lose something by making its frontier models open-weight doesn't hold up If true, that cuts both ways though right? It also means we lose nothing if OpenAI releases a closed weight model.