EDITOR ’ S QUESTION

Let ’ s not forget , the ICT industry has made a significant shift to reduce its emissions – its carbon emissions now stand at around 1.7 % of global emissions – significantly lower than previously forecast . With advancements in technology , we can expect to see a similar trend in how much AI contributes to global emissions .

Ensuring AI models use clean , trusted and reliable data is essential for reducing processing strain and energy consumption . The effectiveness of AI relies on the quality of the data it processes . Clean data – free from bias – enables AI algorithms to operate faster , more efficiently and reduces the time and energy spent training and retraining models . By providing AI models with unbiased data enriched with comprehensive metadata , organisations can ensure their AI models function both optimally , and sustainably .

Ensuring AI models use clean , trusted and reliable data is essential for reducing processing strain and energy consumption .

using less efficient processes and higher-powered end-user hardware .

On the other side , we ’ re also seeing more organisations leverage AI and Generative AI to unlock sustainability use cases . For instance , Large Language Models can process geospatial remote sensing satellite data to predict acute climate-related physical risks like flooding and hurricanes over short periods ( two to four weeks ). They can also forecast chronic risks , such as heat waves and sea-level rise , for longer timelines ( 2050 – 2100 ). These predictive capabilities enable better climate resilience planning and management of transition risks . p

Also , any AI programming and architecture should use leading-edge algorithms , data storage techniques and right-sized computing infrastructure . This ultimately reduces the amount of energy that would be required for similar data processing and analysis

LEVENT ERGIN , GLOBAL CHIEF ESG

SUSTAINABILITY STRATEGIST , INFORMATICA

www . intelligentcio . com INTELLIGENTCIO EUROPE 33