This is fairly complex. I do think it's true. We can't reliably capture what I think of as the qualitative foundations of life. We capture it by compiling particulars.

At face value this leaves room for relativism. You could set any standard and call the outputs "good". Meeting functional goals that preserve life in general are "good".

Respecting reality and reason, over time, reduces error. You will "survive" better in a naturalistic sense if you know more facts.

We do these things to condition our reality to produce a reality where human flourishing is the most likely result.