Will AI replace meteorologists? A private-sector forecaster's view

1 reply

• JK

  Jenny KafkaGiverOct 30, 2026

  First, I don’t think AI is making meteorology jobs obsolete. There is so much more to the job than simply reading a weather forecast and taking it at face value. A big part of being a meteorologist is knowing how to interpret the forecast and, more importantly, understanding the risks and uncertainties around that forecast – and that judgement comes from experience. For example, you might recognize risks associated with certain teleconnection patterns because you’ve seen similar setups many times over the years of forecasting.

  
  

  We now have many AI-based weather forecast models, but NONE of them are 100% accurate. If we ever reached the point where they were, then maybe meteorologists would have more reason to worry!

  
  

  More broadly, I think AI brings pros and cons. One major benefit that I’ve seen is that AI has enabled more non-technical colleagues to experiment with coding and build projects from scratch, from data integration to automation. This can dramatically speed up development.

  
  

  The downside is that problems can arise when someone relies entirely on AI to build something without actually understanding what the code is doing. Maybe the code isn’t extracting or aggregating the data in the exact way you intended. Maybe the code is quietly archiving data somewhere you didn’t realize, and after several months the pipeline fills up your available storage and suddenly other processes start failing. So, I think AI can make many of us more productive, but we also need to understand what is happening beneath the hood, so to speak.