New research has shown that an AI model developed by Google-owned Deepmind has outperformed traditional weather forecasters for the first time.
The peer-reviewed study, published in the journal Science, said that Deepmind's GraphCast AI model "significantly outperformed" the conventional system for predicting the weather, which is run by the European Centre for Medium-range Weather Forecast.
The AI model can produce an accurate 10-day forecast in under a minute and represents "a turning point in weather forecasting," according to the paper.
"Weather prediction is one of the most challenging problems that humanity has been working on for a long, long time," Pushmeet Kohli, Deepmind's VP of research, told Business Insider.
"If you look at what has happened in the last few years with climate change, this is an incredibly important problem," he said.
GraphCast is trained on 39 years of historical weather data from the ECMWF. According to the Financial Times, the research showed that GraphCast's three to 10 day weather predictions were more accurate than conventional forecasts.
But Deepmind warned that the model did have some limitations compared to non-AI forecasts when dealing with the uncertainty inherent in longer-term weather forecasts.
Speaking to the FT, the architects of the project said that GraphCast had been able to predict where Hurricane Lee, a powerful storm that hit North America earlier this year, would make landfall three days before traditional forecasters.
ECMWF machine learning specialist Matthew Chantry told Insider that GraphCast represented "a significant and welcome step forward" for the industry.
He told the FT that once trained on a wider variety of historical data, models such as GraphCast would prove to be much cheaper than current weather forecasting methods, which rely on powerful (and expensive) supercomputers.
"We might be talking about 1,000 times cheaper in terms of energy consumption," he said. "That is a miraculous improvement."
