I am John Joon Young Chung, a second-year Ph.D. student at CROMA Lab, University of Michigan, working with Professor Walter Lasecki. My research area is focused on HCI, crowdsourcing, and AI. I am interested in ways to elicit more complex inputs from humans to make machines more effective and efficient. There is no reason that we should only rely on simple annotations to teach machines because people can do more complex and cooler things! For example, can we use human estimation or prediction to help AI better solve problems? Can we leverage a prototype agent that is created by collective humans to better scaffold AI in the training process? I hope we would be able to build Human-AI systems that can solve a wider set of problems with approaches like these.