Balancing Data Preservation and Practicality in Historical Contexts

I work as a chemist, and my background in data science has me thinking about the ethics of information preservation. It seems like we have the technical ability to save almost everything now, from digitized archives to massive datasets. But does that mean we should? I’m curious about where others draw the line.

For example, in my field, we generate enormous amounts of raw experimental data. Some of it feels trivial, but you never know what future researchers might find valuable. At the same time, storing everything indefinitely has real costs financial, environmental, and even in terms of making the truly important data harder to access. Is there a point where preservation becomes impractical, or even irresponsible? How do you decide what’s worth keeping for the long haul? I’d love to hear perspectives from different fields on this.