Solving Language and Solving AI


This talk focuses on drawing parallels between the problems in artificially generating/understanding natural languages and the problems in general AI tasks. The hypothesis of the talk is that if we can ‘solve’ natural language, we also cover a lot of ground in solving major AI problems. The talk mainly focuses on the main problems in AI - sensing and recognition, understanding and exploration, communication and dialogue, and strategy and hypothesis testing and how similar problems are also encountered when thinking about natural language processing.

Find an overview of the talk here along with slides of the presentation.