My best answer is: yes, and no. AI technologies have a vast ecological footprint, and as models continue to advance, training and deploying them will require the consumption of more and more energy. However, the applications of such technologies are seemingly endless, many of which are already being designed and operated to actually mitigate climate change impacts. Furthermore, strategies can be used to minimize negative effects like building datacenters in places with colder climates and investing in renewable energy. Lastly, as others have commented, these technologies rely on data and algorithmic processes to mimic and aid human cognition. This means that, if we are not careful, the decisions we delegate to AI can be tainted by the biases implicit in the data we feed it, making it yet another reinforcement of socioeconomic disparities. I believe that AI can, and surely will, change everything, and whether this fact is a positive or a negative all depends on how developers address the ethical elephant in the room. I believe the solution is to slow down this exponential growth just long enough to be able to take a step back and ensure more harm is not created than good. Strategize, every step of the way. Invest in the renewable energy sources. Educate the public and professionals about the intentional/mindful use of these models. Program the technology to identify and challenge biases. Before it is too late, we need to ask ourselves: is my use of AI justified? Am I developing this model in an environmentally conscious way?