Artificial intelligence is present in almost every aspect of our daily lives. It is used in search engines, email clients, media platforms, and even food chains around the world. It certainly makes the world a more efficient place, but not without a cost. Job loss is one of them. However, there is one industry that is still trying to fully automate: weather forecasting.
Despite all the advances in technology, artificial intelligence and machine learning, meteorological companies have not yet handed over the keys to their computer counterparts. While AI models, aided by weather satellites such as GOES-16 and 17, are able to predict subtle changes and weather patterns with amazing accuracy, their efficiency drops dramatically in the event of major weather disturbances. Significant climate change, including natural disasters, often has warning signs that are too subtle and small for most models. Even when identified, models cannot always correlate between the initial characteristics of an example watercourse and its potential inlet. According to Wired, this usually requires the trained eye of a seasoned industry expert. The publication claims that based on more than two decades of weather forecast data (collected by the National Oceanic and Atmospheric Administration's (NOAA) Weather Forecast Service)), Humans are among the two most popular national weather forecast models, and they perform better. Global Forecasting System and North American Medium Range Forecasting System. p>
To what extent? Between 20 and 40 per cent. p>
This is not a small gap, and can easily mean the difference between life and death for those on the path to a hurricane, water eruption, or snowstorm with a quick start. Access to accurate weather information as quickly as possible is what makes evacuation (if needed) or on-site shelter recommendations viable. p>
< p> Climate veterans can perform even the smallest details such as subtle changes in atmospheric pressure, wind speed or "available humidity" and higher quality results than their computer counterparts. This is because, in many cases, weather forecast models do not take these criteria seriously. p>
And to be honest, this is not surprising. Computers are getting smarter and smarter by the day, but they still lack what Humans have always had: the ability to assess situations in a broader context. Bringing machines closer to human memory and field awareness requires a great deal of processing power that is not currently widely available. Only a few computers that might be able to do this are currently in development - at least in the US - and climate groups aren't the only ones wanting to use them. p>
In short. Humans are actively replacing alarming speeds in dozens of industries, but for now, the local meteorologist (or at least the group that gave him his data) is safe. p>