Enabling Sustainable AI

Abstract

Deep tech and the world of AI have been rapidly changing the market share and is one of the most focused industries. While opportunities are great, the pit falls maybe likewise. One key element in avoiding pit falls is the need to develop AI with a sustainable approach or building Sustainable AI. AI is dependent on human intelligence and would undergo similar learning cycles to that of humans such as learning, unlearning, and new learning as part of its machine learning ability. This cycle would
ensure it is relevant based on new data and information that is made available, discovered or generated, be it by humans, nature, industry or by AI itself. It would also need an active supply chain, one that is sustainable to ensure appropriate learning and refresh to ensure the improvement continues. The processes applied from design to decline would be key in creating Sustainable AI. To ensure this process is viable and sustainable, principles like design thinking and systems thinking maybe critical right from
the design process, through the steps carried out in the AI build and its continued maintenance to ensure a more humane and wholistic development of AI. This design approach needs to be profitable, inclusive and circular to ensure stakeholder objectives, human social factors, economic requirements and environmental impacts are addressed.
Keywords: Sustainability, AI, Machine Learning, Artificial Intelligence, environmental, societal, economy, design, Sustainable AI.