Cursor just announced a 1.5 trillion parameter model trained on 100,000 GPUs. Their CEO said it's "as big as Claude Opus and GPT-5.5." That one comparison just revealed something the frontier labs would rather have kept quiet.Parameters are what a neural network learns during training, the internal settings that shape how the model thinks and what it knows. GPT-3, which shocked the world in 2020, had 175 billion of them. We are now at 1.5 trillion.If Cursor's model is the same size as Claude Opu