The increasing volume, resolution and variety of geoscience data, along with increasing capabilities of high performance computing and new machine learning (ML) techniques are providing novel opportunities. Deep learning (DL) is now being applied to track weather events, localise, detect and classify extreme events, or emulate physics models. Therefore, it is timely to examine the value of DL/ML or hybrid approaches, over traditional physics methods. The artificial intelligence (AI) community also uses different criteria to benchmark progress through global challenges such as ImageNet versus how geoscience tackles planetary challenges, explainability and reproducibility of results. As new waves of intelligence are applied to the Earth, whist we face existential risks from exceeding its sustainable boundaries -- what are the emerging opportunities and challenges at the crossroads of this planetary transition, into this new age of ‘Planetary Intelligence’? The panel will explore this question across three topics: Physics-guided ML; EnviroNet & Climate change-AI.
![](images/flag.png)
![](images/blue_ribbon.png)