Author: Mark Coeckelbergh
Publisher: AI and Ethics
Publication Year: 2021
Summary: The following piece discusses how artificial intelligence (AI) can be used to help lessen climate change but can also exacerbate challenges. Examples mentioned are that computer data processing uses energy, can be used by fossil fuel industries, or contribute to excessive consumer consumption that harms the environment. This consideration was something not commonly seen addressed in data science before, and provides a unique perspective. One interesting study that is cited states that training a large natural language processing (NLP) model can lead to 300,000 kg of emissions of CO2. However, tech companies also use AI to help climate change slow down and to benefit the environment generally, so there is a push and pull, but which will outweigh the other is the key question. The article shares several ways to lower the cost to the environment, such as transparency about environmental impacts. The article also discusses the ethics of crafting policy around this issue and individual freedoms vs paternalism.