Mitchell-Olds and Shaw pointed this out in the context of estimating selection gradients years ago....

yep, and how often does it need to be pointed out that without randomization and replication, those cov terms cannot be safely assumed away to zero.....

💯

If one my technical papers on RNAseq or natural selection "resonates" or "aligns" with your perspectives as a 20-something year old undergraduate, I suspect its AI generated word salad.

Better to simply write "I'm really interested in natural selection. Do you have students working on it now?"

I would prefer brief, sincere, and open-ended to word salad b.s. So, "Your research on X is really interesting to me. How can I learn more? What could a student do next?" is 10,000x better than word salad.

On Halloween night, I usually give huge handfuls to really little kids, saying "I'm giving you a ton because your parents are going to steal a lot of it!"