The world is currently facing an unprecedented environmental crisis, and as such, the need for sustainable and innovative technologies has become more pressing than ever. One technology that holds immense promise and concern in equal measure is Artificial Intelligence (AI).
You might be surprised to learn that I did not write the opening sentence of this blog article. It was actually generated by the new AI language model ChatGPT.
Since its launch in November 2022, ChatGPT has revolutionized the way many of us are using the internet. With its ability to produce human-like responses to natural language input, along with its capacity to carry out a wide range of tasks, it has gained both popularity and notoriety in just a short space of time.
While some view it as a useful tool to enhance productivity and improve access to information, others have been quick to warn of the dangers the technology could pose to society, citing everything from the perpetuation of misinformation and cultural biases to its potential to cause mass global unemployment through increased automation.
However, there is a less discussed but equally important issue surrounding AI: its environmental impact.
AI’s Carbon Footprint
According to a recent report by researchers at Stanford University, the amount of energy used to train GPT-3, the language model behind ChatGPT, amounted to almost 1,300 megawatt hours of power, producing the equivalent of over 500 tonnes of carbon dioxide emissions. To put this in perspective, that is roughly the same as the emissions produced from driving 1.65 million miles in an average sized petrol car – that’s 66 times around the Earth’s equator!
And this only accounts for the emissions used to train the language model. A study of the AI language model BLOOM estimated that, when factoring in other processes such as equipment manufacturing and energy-based operational consumption, the overall emissions more than doubled. While further research is needed, there is no doubt that AI’s carbon footprint is a serious concern that needs to be addressed.
While AI’s impact on the environment may pale in comparison to other areas of the tech sector, which collectively make up around 2 to 3 percent of global greenhouse gas emissions, this revolutionary technology is only likely to grow both in use and complexity in the next few years. With this growth, there is likely to come an increased need for energy, and as long as that energy is generated from largely non-renewable sources, a rise in greenhouse gas emissions is regrettably inevitable.
However, there is a silver lining. AI could in fact become a force for positive change in the fight against climate change.
AI as a Sustainable Tool
One example of how artificial intelligence is already being used to reduce greenhouse gas emissions is through automated traffic systems which collect real-time data on traffic patterns. These patterns are carefully analysed using machine learning algorithms to make predictions and optimise traffic flow. With the obvious financial, productivity, and environmental incentives, the uptake of transport optimization is set to expand massively in the next few years and is likely to become a common feature in many cities across the globe.
According to a report from PwC, this optimization of current transport processes, along with the adoption of AI technology which is already taking place in the energy, agriculture, and industrial industries, has the potential to reduce global greenhouse gas emissions by as much as 2.4 gigatons by 2030 – this is around the same as the annual carbon footprints of Japan, South Korea, and Germany combined.
Of course, there are many factors that could impact these estimates, such as the availability of accurate data and the speed at which these emerging machine learning technologies are adopted. However, even by the most conservative estimates, AI will almost certainly offer at least some form of positive contribution in reducing our collective greenhouse gas emissions.
The big question is whether AI will ultimately contribute more to the reduction of emissions than it generates through the energy used to train, power, and store the data the technology is responsible for producing. The answer will depend largely on how quickly the tech industry can transition away from the widescale use of fossil fuels and move towards a future where the use of renewable energy is the rule, not the exception.
As no one can be truly sure whether or not this transition will be possible to achieve before the damage is irreversible, I decided to ask ChatGPT how it thought the situation may play out. This was its response:
“As an AI language model, I don’t have personal opinions, but I can provide an objective answer based on the current state of scientific research and industry trends. According to experts in the field of renewable energy and climate change, it is technically feasible for humanity to make the transition to renewable energy in time to avert the climate crisis, but it will require significant political will, investment, and societal changes to achieve this goal. The technology and resources for renewable energy are available and advancing, but it will require concerted efforts from governments, businesses, and individuals around the world to make the transition quickly and effectively.”
…I couldn’t have said it better myself.
It is clear that artificial intelligence has the potential to both contribute to and mitigate the effects of climate change. While it is important to acknowledge and address the carbon footprint of AI, it is equally important to recognize the ways in which it can help reduce greenhouse gas emissions and protect our environment.
The key is to continue to develop and utilize AI in a responsible and sustainable way, while also prioritising the transition to renewable energy sources. With the right approach, we can harness the power of AI to create a more sustainable and equitable future for everyone.