The following article aligns with our discussion of using AI technology to create deep fakes. However, this article identifies an experiment conducted by a group of University of Chicago researchers. The experiment involves using computer algorithms to create fake Yelp restaurant evaluations. The researchers used a machine-learning technique known as deep learning to analyze letter and word patterns used in millions of existing Yelp reviews. Usually when there are mass postings of fake reviews, for example when Hilary Clinton’s book was launched, websites are able to identify it. However, with the researchers’ use of AI, Yelp’s filtering software had difficulty spotting many of the fakes. Even human test subjects were unable to tell the difference between real and fake reviews.
The researchers wanted to conduct this experiment in order to see whether it was possible for individuals to make fake reviews on a mass level in order to impact people’s use of a product, book, restaurant or anything else. This interest displays how things focused on in the media like “fake news” or “deep fakes” really can impact the research and developments in the communication field.
For now, the approach to create fake critics remains to be “crowdturfing” where mass amounts of people are paid to comment on websites. “In crowdturfing online reviews, an attacker creates a project on the Mechanical Turk site and offers to pay large numbers of people to set up accounts on Amazon, Yelp, TripAdvisor or other sites and to then post reviews intended to either raise or sink a product or service’s money-making prospects”.
From here the researchers plan on looking at ways to detect fake news – currently its difficult to create fake news with AI because computers cannot generate the “human touch” in articles however if AI can create fake reviews – anything is possible.