wouldn't it be more sustainable if

I work at Google, not directly on AI, but... when you are working under multiple layers of middle management, there is an inordinate demand for pieces of paper that say "obvious" things on them (and have at least a veneer of credibility). It turns out that, while many of these things are "obvious" to you and I, they are much less obvious to people who work at a significantly higher layer of abstraction than we do.

I do not have any special insight into what my company plans to do in the future, and I do not want a hypothetical to be misinterpreted as a veiled leak, so I will not be able to be any more specific than that. But I will point out that, in general, middle and upper management are much more likely to act on something if they have seen a (credible) piece of paper explaining it and giving reasons to believe that it is true. This goes for both "obvious" and "non-obvious" things.