A.I. Could End Fake News
By Alan Nero
Could AI be the secret weapon society needs to pull free from the suffocating mire of the “Fake News” era?
“Fake News” is indeed a legitimate issue in Western society and its foundations are found on social media. Not to be confused with the false narrative pushed by both public officials and private citizens alike who muddle the difference between inconvenient news and actual false reporting.
While I’ve previously outlined the problematic behavior exhibited by those who contribute to this false narrative, we face a greater epidemic of baseless claims, inaccurate reports, and sensationalist headlines. The results of which grow increasingly severe with the reputation and influence of those who propagate these lies. Worse still, the irresponsible actions of social media platforms only help fan the flames and disseminate “Fake News” to the public.
One of the most disturbing instances of the spread of “Fake News” through social media is the escalating conflict between Buddhists and Muslims in Myanmar, fueled by Facebook. Where Muslims are regularly targeted by hate speech and lies, which lead to violent action.
The greatest difficulty faced by those who wish to fight the swell of “Fake News” is the staggering amount produced and shared on a minute-to-minute basis in the age of global-communication. A 2018 Forbes article stated there are 1.5 billion active users on Facebook alone every day. It is logistically impossible for even a large dedicated force of auditors to sift through the over 400 million posts made daily on only one platform. Let alone fact check each article, video, and post shared across the web.
However, promising advancements in Artificial Intelligence could be the solution to this dilemma. According to AP’s The Future of Augmented Journalism, Supervised Machine Learning is an AI system which, when supplied with labeled examples as an input, is able classify new information as an output. Essentially, an AI system can be fed examples of both verified and falsified articles, learn how to identify each type of article, and then differentiate between future true and false articles. Furthermore, the nature of such a system turns the formerly daunting statistics into an incredible advantage. As the authors of the AP article stated,
“The more data a team has to teach its system, and the more accurate that data, the less likely the machine is to commit any errors.
No machine-learning algorithm is going to be 100 percent accurate.”
While such a process would not lead to AI being solely responsible for automatically verifying all reports and claims on the web, it would make the act of isolating and targeting all future “Fake News” not only possible, but easy. At that point, dedicated auditors could then focus their energies and fact checking and disproving false claims and inaccurate statements.
The joint efforts of AI Supervised Machine Learning and devoted researchers could put an end to the dangers of “Fake News”.
