Large language models, scaling laws, AI Alignment, democratization of DL