Does a PHD thesis in Machine Learning even make sense, given how fast the field develops

After finishing my masters degree in Data Science, I would love to get into ML-Research. I know that the majority of research positions do require a PHD so I am thinking about pursuing one after graduation.

Now what I am concerned about is that I start to write my dissertation on something which builds on the state of the art but will eventually become completely irrelevant after a year or two because the state of the art has progressed massively …

data data science datascience love machine machine learning phd research science sense thesis thinking