Not sure if this is really “code red” for Google. First of all they also have strong language models (PaLM) but more importantly they already have a huge user base. Even though chatGPT achieved a million active users within five days (which is already impressive) Google would probably have a few orders of magnitude more right on day one if they were to ship such a feature. One reason Google might not want to do it right know, is that they need to care more about public perception. Shipping an engine that obviously does mistakes seems way more dangerous to a corporation like Google than a “startup” like OpenAi when it comes to public perception. What do you think?