Google had done significant research in language models but did not achieve public recognition equivalent to OpenAI's ChatGPT because Google lacked the key innovation of reinforcement learning from human feedback (RLHF), which ChatGPT pioneered.