Implicit Costs of AI: Environmental Impact

There is no secret that artificial intelligence is a resource-hungry technology due to the overwhelmingly exorbitant amounts of energy required to ensure the sustainable running of supercomputers. The ones, undoubtedly, are needed to store and process the enormous volume of content for machinery learning. According to the <a href="https://arxiv.org/abs/1906.02243" rel="noopener ugc nofollow" target="_blank">research</a> carried out by scholars at the University of Massachusetts, the training of a common AI system can emit over 626,000 pounds (!) of carbon dioxide equivalent, whiсh is nearly comparable to the life cycles of 5 cars. Indeed, the global concern regarding redundant energy consumption is not confined to the AI data centres, but it would be fundamentally wrong to ignore their indispensable contribution for the appalling climate alterations. And apparently, the ball began to roll. To be more specific, the issue began so remarkably that AI engineers were forced to tackle the drawbacks and rethink of the sources powering dara farms. <a href="https://medium.com/predict/implicit-costs-of-ai-environmental-impact-5e38d8c86492">Click Here</a>