Tag
A commentary arguing OpenAI is likely ~6 months ahead of Anthropic internally, deliberately withholding its frontier models and releasing only what it's comfortable showing to preserve its compute and strategic advantage.