AI predicts 2.5B coronavirus infections

An AI simulation model predicts catastrophic effects of the novel coronavirus that originated in Wuhan, China, but the credibility of AI models is being called into question.

As many as 2.5 billion people could be infected with the disease within 45 days, with 52.9 million deaths, according to the simulation. Fintech startup founder of HedgeChatter, James Ross, built the AI simulation to estimate the global impact of the virus, prompted by another site that is tracking coronavirus globally.

As of publication, the virus has claimed 565 lives and infected more than 28,000. So far, Ross’ model has actually been accurate in predicting the following day’s publicly-released data within 3%, Forbes reported.

However, the model doesn’t account for every factor, including improvement in the reaction to the virus and conditions. In addition, the 2% mortality rate for coronavirus, which is extremely high and has prompted concern worldwide, could be wrong if some infected persons are asymptomatic. And it is unlikely to be as deadly as Ross’ AI model predicts.

See the full story below: