MIT EECS Undergraduate Research and Innovation Scholar
Artificial Methods to Find Precedents Using the Genesis System
Patrick H. Winston
This project will take the goals of the original Genesis project to a new level by finding precedents in a scenario to strengthen causal relations. This will help to fill in details of “why” something happened, and to understand motives for complex problems. The Genesis group developed a theory called the “Strong Story Hypothesis,” which asserts that a deep understanding of stories is a vital component in human intelligence. I propose that an equally important component of intelligence is the ability to form hypotheses regarding the future by analyzing precedents that caused similar events to occur in the past. In this research, we will analyze the outcomes that resulted from actions and events, and map them to similar outcomes that could result from the current scenario.
My motivation for this project came from my exploration of the topic of Artificial Intelligence in Patrick Winston’s 6.034 and 6.833 classes. In 6.833, I performed my final project on the topic of integrating metaphors into machine conversation. My internship at Microsoft reinforced my passions, and I seek to explore AI further in the Genesis Group.