Looking back at the CCAI summer school 2026

While Europe was simmering in the summer heat, it was winter in South Africa and thereby primetime to do indoor activities, like attending a virtual summer school on AI for climate change—which I did. The MOOC was offered by Climate Change AI and while the minimum requirements probably would take about the equivalent of a week at most, I ‘overconsumed’ with more foundational and sectorial modules (8/6 and 10/4, respectively), which consisted of online lectures, tutorials, quizzes, evaluation forms, and additional readings, which took at least an estimated 2 weeks full-time equivalent, and counting. It was most certainly value for money for the amount and quality of materials provided. The computing contents of the modules were heavily skewed towards the subsymbolic AI, and so yes, I did get to play with CNNs, GNNs, reinforcement learning, random forests, regression, decision trees, LSTM, hyperparameter tuning, LLMs and so on, in Google Colab at that. Technologies used included Python,…