The most popular and comprehensive Open Source ECM platform
Let us know more
Artificial Intelligence: AI’s Carbon Footprint
The popularity of Artificial Intelligence and the massive computational resources required by AI create an industry with a mammoth carbon footprint. The computational resources for “best-in-class” AI models are doubling every 3.4 months which equals a more than 300,000 times increase since 2012.
Dan Simion, vice president at CapGemini, said that “if you’re going to run AI, you’re going to need ‘the machines,’ so you’re going to leave a carbon footprint. As you get more data and the models become more complicated, the more energy you’re going to consume to get the AI models you’re looking for.”
Dan Jurafsky, CS professor at Stanford, said that “There’s a big push to scale up machine learning to solve bigger and bigger problems, using more compute power and more data. As that happens, we have to be mindful of whether the benefits of these heavy-compute models are worth the cost of the impact on the environment.”
Donna Lu, a contributor at New Scientist, wrote that “training artificial intelligence is an energy-intensive process. New estimates suggest that the carbon footprint of training a single AI is as much as 284 tons of carbon dioxide equivalent — five times the lifetime emissions of an average car.”
Research at Google suggests the following “4M” best practices to reduce the AI carbon footprint:
- Model – Choose an efficient model, like a ‘sparse model’.
- Machine – Use AI-optimized processors rather than general-purpose processors.
- Mechanization – Computing in the cloud rather than on-premise tends to use high-volume energy-efficient data centers rather than smaller data centers with older equipment.
- Map Optimization – Select cloud processing locations that have the cleanest energy footprint.
