> We plan to publish one final academic paper later this year that will detail the extensive set of improvements we made to the algorithms’ efficiency and poten

> We plan to publish one final academic paper later this year that will detail the extensive set of improvements we made to the algorithms’ efficiency and potential to be generalised across a broader set of problems. I'm fascinated to see what the next step for this AI is. Anyone care to speculate what a system like this could most readily be applied to?