Well, I'd politely agree to disagree. Google scale is defined by novel, radical approaches, like for example inventing map/reduce, writing papers on LLMs others

Well, I'd politely agree to disagree. Google scale is defined by novel, radical approaches, like for example inventing map/reduce, writing papers on LLMs others then implement successfully, or creating something like Kubernetes. The specific topic here is not one of those google problems to me, as I can compare it to other problems we already solved. But yes, we could miss that critical point where a totally different problem domain emerges just from one order of magnitude more, so fair game.