Google DeepMind has launched a new, allegedly revolutionary artificial intelligence tool named GenCast, which supposedly creates very accurate 15-day weather forecasts faster than with traditional methods. The company made this claim in an article published on Wednesday in the Nature Journal, stating that it will possibly beat current worldwide norms, particularly in tracing devastating storms and saving human life.

GenCast is, according to Ilan Price, the lead author of the study and a senior research scientist at DeepMind, faster, and better than previous methods. "And it's more accurate," he added, pointing out the program's potential in redefining weather forecasting.

DeepMind's latest achievement follows the earlier success of its AI weather program launched in late 2023, which provided reliable 10-day forecasts. GenCast, however, goes several steps beyond that in terms of both speed and accuracy.

"I'm a little bit reluctant to say it, but it's like we've made decades' worth of improvements in one year. We're seeing really, really rapid progress," said co-author Rémi Lam, lead scientist on the previous project.

How does GenCast compare to global standards?

The European Center for Medium-Range Weather Forecasts generally considered the gold standard for atmospheric prediction, is accessed by 35 countries throughout the world. But in strict testing, GenCast scored higher than ECMWF's Ensemble Prediction System 97.2% of the time.

The study tested the accuracy of both systems over 15 days by analyzing 1,320 global wind speeds, temperatures, and other atmospheric metrics. The results indicate that GenCast could be a precursor to major breakthroughs in operational weather forecasting.

What experts say about GenCast

Atmospheric scientist Kerry Emanuel, known for his work on atmospheric science, called the achievement a “big deal,” describing it as an important step forward. Emanuel and other experts have previously pointed out that extending reliable forecasts from 10 days to 15 days would have major socioeconomic benefits, as such advancements could help communities better prepare for extreme weather events.

DeepMind's GenCast could bring a transformative change in weather prediction, making it possible to have more accurate planning and response for natural disasters. The team believes that this tool could “open the next chapter in operational weather forecasting,” giving hope that it could mitigate the worst impacts of extreme weather on vulnerable populations.

By combining speed and accuracy, GenCast shows how AI can change the world of industries, saving actual lives in the process.