You should rephrase the title to "young scientist begins learning about search engines". It's not the first time I've been teased by the promise of improved search and I'm always disappointed with the realization how infinitesimal the odds of unseating Google are. As such, anything worth mentioning is going to come from a PHD - not a kid who took an AI class a year early (I don't know anyone moderately interested in computer science who hasn't heard of Markov chains). Although they didn't mention anything specific about his algorithm in the video of which I'm still curious in seeing, I'm certain it's not any faster or more relevant than any other other basic search (not to be a debbie downer, there's simply smarter minds who've worked far longer on the same problem).
I actually have a bot that has been doing this for several years. It stores word relations and keeps track of what words are used with others. It then can rank potential words that it has seen from your query before and use those as a subject to build a sentence from or as a highly targeted query for searching.
Seventeen-year-old Nicholas Schiefer is on his way to developing the next Google for the social-media world. The Intel Foundation Young Scientist Award winner sits down with digits to explain.
Sounds like an interesting idea; I hope he publishes more details. I also hope the next interviewer is not so detached.