The Advantage of Coming First in Machine Learning Cumulative

Is the advantage of being first in machine learning cumulative? Or is it just transitory, and that it only lasts for a certain while. I think some of the advantage is premanent, in the sense that if you have a good product and people start using it, then you have a lot of iews. So i think in some aspects, this first move or advantage is extremely important because you have more data to learn from. There is a race to develop better machine learning algrithem. And there is certainly an advantage to coming first. Having said that, there are also lots of opportunities for those who are just starting, for example.