AI system training results in high energy and carbon footprint expenditures. One AI transformer, for instance, consumes five times as much CO2 in power as a gasoline car does over the course of its lifetime.

The number of operations completed per second per watt used, or TOPS/W, as a measure of computing performance, has led to the development of organic integrated circuits, or simply organic chips, as a potential answer. Organic chips, however, have not yet enhanced the real front-end learning and machine training process, simply because they are semi-sentient and can have fluctuating "motivations" and "morale" while having shown improved basic functions in machine intelligence utilised for data classification.