phycodurus

that's not really how references work

It was when I learned that the genetic code actually resides in the set of tRNA-synthatase enzymes that couple the appropriate amino acid to appropriate tRNA.

From the last paragraph (of the paper): "The overpotential ... probably precludes economic viability for this catalyst..." There's a reason this is published where it is (and not Science or Nature or JACS). Popular Mechanics turned this reasonable research into click bait.

cat <that article> | sed -e s/data/your data/g

I wish someone would build an interactive teaching AI, (perhaps in a mathematical context at first). For example, the user might start with a goal such as, "I'd like to understand singular value decomposition." The AI would interactively assess the user's level of background and begin instruction at the appropriate level, leading to the desired goal.