From Visual Question Answering to multimodal learning: an interview with Aishwarya Agrawal