Got truthiness? Why AI bots will become our new arbiter of news
Carmen Hughes
The New York Times "Run-Up" podcast recently examined the shifting tides in media reporting and readers, viewers, and listeners. Audiences have increasingly become distrustful of news articles and the organizations behind them. The tribalization of the media has divided audiences largely based on their biases. We've reached an acute level where everyone questions everything, making truth absolutely relative.
Unfortunately, fact-checking seems to have gone largely by the wayside. Indeed the economic realities of speed, clicks, eyeballs, and share of audience seem to have trumped accuracy and substance. Everywhere you turn, questionable stories or completely fabricated conspiracies abound with little to no substantiated facts.
If we have no trusted sources or gatekeepers, how can audiences find factual stories and reporting amid the wild falsehoods? If the media is indeed losing its role as ‘the arbiter of truth,” where can we turn for a reality check?
Perhaps artificial intelligence (AI) holds the promise in becoming our new arbiter of truth?
A number of young AI companies are making great strides in computer-generated news reporting. For instance, Narrative Science teaches computers how to write journalism. The company marries advanced pattern recognition software with natural language generation to create algorithms that resemble an actual writer. Its founders predict that in less than 15 years, or by 2030, 90% of journalism will be written by a computer.
During the Rio Olympics, The Washington Post used an AI-based, robot “reporter” to successfully prepare and post hundreds of sports updates. Earlier this year, the Associated Press began using Automated Insights' platform called Wordsmith to generate content. Wordsmith acts as a personal data scientist and can sift through reams of data to generate corporate quarterly earnings reports. Wordsmith also produces daily Fantasy Football game summaries for millions of Yahoo users during Fantasy Football season.
So the idea of a smart, non-partisan, non-emotionally invested, AI-based computer robot could serve as the new backstop for factual reporting and unbiased fact-checking. AI-based computer robots would bring a refreshingly unbiased lens. Undoubtedly, it would lack the creative flair and talents of a veteran journalist. However, the AI robot reporter would be able to drill down into publicly available data to set any record straight without any emotional bias.
In the near future, we will undoubtedly have Watson, Siri and Alexa-like resources that can serve as our newly crowned arbitrator of truthiness. And yes, different factions will always exist that condemn these AI computer writers and their findings or reporting merely because the articles don’t align with their personal bias. To preempt this, open sourcing the algorithms and code should be a consideration.
In Kevin Kelly’s new book, “The Inevitable”, he introduces the notion of a "centaur", which is a mixed team of AI and humans. Centaurs have become the best chess players in the world, blending their skills to become unbeatable. Big media organizations can also benefit by implementing a centaur-based approach that doesn’t do away with human talent but uses complementary skills to produce and distribute less biased, more accurate news reporting with full fact-checking.
AI bots are in the initial stages of taking over our newsrooms (as well as a broad spectrum of other job professions). Perhaps many will welcome the rise of the machines that will eventually replace today’s echo chambers and biased news organizations that have evolved, where bite-sized nuggets of truthiness run rampant. Ideally, we can see the rise of the centaurs to bring about a return to verifiable, unbiased reporting.