Today I presented my new work for the first time. It was surprisingly easy to talk about, even though my understanding of the subject matter is not perfect.

The topic is "predicting stability of solid crystals using graph neural networks". Far more exciting than any thermodynamics I used to do.