Tabular data has been said to be the final frontier of machine learning. This might be a bit overdramatic, but there is some truth to it; despite rapid advancements in AI, the models we use to analyze tables of data has remained the same for around 30 years.
I've been working with a team of independent researchers to try to push the state of the art of tabular data forward. We plan on publishing, presenting, and innovating. It's all very loose right now, we're still trying to get our brains around the issues of tabular modeling and useful goals worth approaching. By donating, you allow me to allocate more of my time towards this project.