We always knew that one day, we’d be bowing down to our robot/artificial-intelligence overlords, and while today isn’t quite that day — we’re definitely getting closer.
A new machine learning program, that yes, is AI (artificial intelligence) can identify COVID-19 conspiracy theories on social media — while these theories are developing.
This is great news for public health officials who have spent the pandemic battling to combat misinformation online while doctors have been working to debunk COVID-19 myths for months.
The algorithm, which was developed by Los Alamos National Laboratory researchers, not only identifies COVID-19-related conspiracy theories — which always gain their strongest foothold through social media — it also models exactly how they evolve over time.
“Thought I’d Share First“, the title of the study, used publicly available anonymised Twitter data, specifically looking at four COVID-19 conspiracy theory themes. In case you were wondering, one of them involved 5G.
The other three themes, to the surprise of absolutely no one who spends any time on the internet, included the theory that the Bill and Melinda Gates Foundation engineered, or had otherwise “malicious intent” related to the virus; that COVID was bioengineered or developed in a laboratory; and that the COVID-19 vaccines would be dangerous.
On that note, here’s how to talk to a friend who’s on the fence about the vaccine.
If you’re intrigued as to how they then developed the algorithm that could identify these types of conspiracy theory tweets on their own? The research team identified subsets (from a body of data that originally consisted of approximately 1.8 million tweets) that matched the four conspiracy theories using pattern filtering — then they hand labelled several hundred tweets in each conspiracy theory category to construct training sets.
One thing the team found? Unsurprisingly, that misinformation Tweets contain more negative sentiment than their factual counterparts — and, that conspiracy theories evolve over time. In fact, they incorporate details from unrelated conspiracy theories, as well as real-world events.
“It’s important for public health officials to know how conspiracy theories are evolving and gaining traction over time,” said co-author of the study, Courtney Shelley.
Public health officials can use this to explore changes in word importance among topics within each theory — we know how easily words from health professionals can be misconstrued by conspiracy theorists. See: any word Bill Gates has ever spoken (microchips).