The chaos surrounding the new coronavirus pandemic, which causes Covid-19, claimed an unlikely victim: Artificial Intelligence (AI) algorithms.
The pandemic is confusing the machine learning algorithms that are programmed to make sense of our online behavior. Human behavior has changed since people began to isolate themselves in their homes to try to stop the virus from spreading, and algorithms are having trouble understanding this new pattern of behavior.
According to MIT Technology Review, algorithms that recommend products on Amazon, for example, are having trouble understanding our new lifestyle. Although machine learning tools are built to receive new data, they are generally not as robust to adapt so dramatically in a short period of time.
The MIT Technology Review reports the case where a company that detects credit card fraud had to intervene and adjust its algorithm to explain a wave of interest in gardening equipment and power tools. In another case, an online seller found that his artificial intelligence was ordering products that no longer matched what was being sold.
There is also the case of a company whose artificial intelligence, used to recommend investments based on the analysis of news sentiment, was confused by the generally negative tone in all media.
“The situation is so volatile,” said Rael Cline, CEO of the algorithmic marketing consulting firm Nozzle, according to Futurism. “You’re trying to optimize for toilet paper last week, and this week everyone wants to buy puzzles or gym equipment.”
While some companies are devoting more time and resources to manually target their algorithms, others see this as an opportunity to create better machine learning models.