- Start with a simple map (for both materials discovery and materials synthesis)
- Understand the rough purpose of each branch
Example:
- Why does DFT exist? What problem does it solve? Where does it fit in the workflow? What are its limitations?
- What bottleneck does this remove?
Examples:
- DFT ā predicts properties without synthesizing every candidate.
- High-throughput experimentation ā generates data faster.
- Lab automation ā scales experimentation.
- Predictive synthesis ā bridges the gap between a promising material and actually making it.
- Foundation models ā search larger design spaces.
- Once I understand the bottleneck each branch addresses, I can evaluate which bottlenecks feel important, unsolved, and interesting to me.
Pass 1: Map
Spend a few hours to a few days understanding:
- purpose
- inputs
- outputs
- limitations
- relationship to neighboring fields
Pass 2: Deep dive
Only for branches that satisfy:
- intellectually interesting
- strategically important
- likely to matter for the kind of work you want to do
Everything else stays as a box on the map.
The mistake isn't having incomplete knowledge. The mistake is refusing to leave boxes black-boxed.
A good mental map often looks like:
I know 10% of 100 things and 80% of 3 things.
vasa
