Slashdot Mirror
AI Bookworms Seek To Predict Human Behavior (thestack.com)
An anonymous reader writes: Creating virtual assistants that can understand and anticipate human behavior and needs is one of the current lodestars of artificial intelligence research. Now, researchers at Stanford University have decided to approach the problem by using descriptions of everyday human activities found in online fiction, namely 600,000 stories from 500,000 writers at online writing community WattPad – input totalling 1.8 billion words – to inform a new knowledge base called Augur, designed to power vector machines in making predictions about what an individual user might be about to do, or want to do next. The scientists suggest that crowdsourcing or similar user-feedback systems would likely be necessary to amend some of the more dramatic associations that certain objects or situations might inspire. As the research notes, 'If fiction were truly representative of our lives, we might be constantly drawing swords and kissing in the rain.'
basically take a large sample of alleged interactions of humans as imagined by a particular group of writers, make statistical analysis of them, make predictions based numbers and averages spewed out.
in other words, average of prejudices of writers who chose to use a website, is going to to be, so called artificial intelligence's, idea of human behavior.
oh my! how artificial! how the science has advanced!
-
btw just imagine this.
how will so called artificial intelligence do if while playing chess, it played the move that average human will take(as derived from all human player games as stored in chess databases) in a given position .