This C&EN article <Is machine learning overhyped?> is great in its candour and depth about how machine learning is affecting chemistry. As chemistry is the central science, some points also definitely apply to science at large.
Here are some of my takeaways:
- “30% of respondents use machine learning regularly in their work.”
- QSAR models have been using machine learning all along
- Chemists (and other scientists) need basic coding skills
tl;dr - Machine learning is kinda overhyped but you still need to keep up with it
I am taking this to heart: I am actively learning Python now since the beginning of the year and am attending some introductory machine learning lectures. Thankfully my brief encounter with Pascal in secondary school and all my Bash scripting and CHARMM scripting during my PhD make learning Python somewhat easier. Can’t say the same for machine learning though, the learning curve is definitely steeper for me…