Tom Pashby reports on some of the benefits and applications of AI in environmental management.
Every part of our lives is increasingly touched by technology, and every sector is at least experimenting with, if not fully adopting, emerging technologies such as artificial intelligence (AI).
Some technological purists would say we have not yet developed an AI because neither ChatGPT or any other large language model has been able to successfully emulate humans, or go beyond human intelligence, a step that would take us towards achieving what is known as artificial general intelligence (AGI).
However, plenty of credible organisations are promoting their so-called ‘AI’ products or the ways they have used AI to optimise their operations, and this is having an increasing impact on jobs and skills.
Environmental management and assessment is a particularly interesting area for the adoption of AI. In part, that’s because controversies associated with AI include its environmental impacts. It is also because of the increasing digitalisation and commodification of our environment, which, somewhat invertedly, has primed the sector for advanced analytics.
Blockchain & Climate Institute Director General Alastair Marke advocates for the widespread adoption of AI across the climate action movement.
Marke said: “The use of AI can help us understand climate change impacts a lot more efficiently even at a very granular level.
“The use of blockchain or distributed ledger technology can help implement climate actions, particularly around renewable energy trading, carbon markets, carbon credit trading, the disbursement of climate finance from developed to developing countries, or from investors to the project developers.”
Benefits and applications of AI in environmental management
Large organisations in the environmental management sector, employing thousands of people, are adopting AI at pace.
The United Nations Environment Programme (UNEP) Chief of Digital Office Sally Radwan, told the Green Careers Hub her organisation is taking a balanced approach, driving forward adoption while recognising potential negative impacts.
Radwan said: “AI offers transformative opportunities for the environment, including climate change mitigation, nature conservation and restoration, and pollution reduction.
“However, it is essential to consider both the potential benefits and the potential drawbacks of AI. At UNEP, we consider two sides of the digital coin: digital sustainability and sustainable digitalisation.”
Radwan said UNEP has been “researching the ethical and effective use of large language models (LLMs) for environmental context through an application – EnvironmentGPT.
“Our objective for EnvironmentGPT is to create a reliable digital resource for environmental management and decision-making, based on the world’s environmental science.”
She was keen to reassure that EnvironmentGPT is “trained with UNEP-vetted data” and “ensures the highest standards of integrity, quality, and trust.”
SUEZ recycling and recovery UK Chief Sustainability and External Affairs Officer Adam Read said: “At SUEZ we’ve been exploring the potential of AI at several of our sites to enhance our operations, whilst working with a number of partners on the journey.
“One of the uses for AI that we’ve been utilising for the last 12 months or so is around target material identification on the picking belts at our materials recovery facilities (MRFs) and identifying these to the picking teams to increase capture and quality of the materials we source for re-processors. “We are looking at expanding the use of AI technology at our MRFs to standardise the process and provide more accurate and efficient data.”
Material costs and wider environmental concerns
University of Reading researcher Simon Driscoll said: “There are definitely concerns about the energy cost and environmental impact of AI, which people are increasingly aware of.”
He said academic and other professional work is ongoing to highlight and understand “serious concerns about the energy consumption of AI, for example large language models.”
Driscoll went on to say: “Aside from energy consumption, LLMs require latest generation hardware”, pointing out that this includes the mining of rare-earth metals and the abstraction of water.
However, he also said: “AI may bring many positives and more efficient processes and more novel applications that do a lot of good. It is a case of positives and negatives, not one or the other.”
Cardiff University Research Associate Jedrzej Niklas said he works with forestry professionals who are using drones to collect data and use AI to provide novel insights.
Niklas said: “The foresters are spending less and less time in the field. They’re using more computers and more drones to look at the forest from above. They know their land less and less.”
Companies are anxious to appear with or ahead of the curve, to impress and reassure clients that they are getting the best and most cost-efficient services, creating new jobs in the process. This means a significant minority of workers in environmental management roles are likely to need new skills to understand and use AI to enhance the work they do.
However, AI has handed a new lever of power to employers, with employees being put at risk of being replaced by AI tools.
Several of the academics and practitioners investigating and using AI raised concerns about the regulatory framework within which AI is deployed, or the lack thereof.
In the 17 July 2024 King’s Speech, the new Labour government promised legislation that would cover AI, but changes to the regulatory system are notoriously slow, and technologies like AI are known for developing exceedingly quickly.
This is a guest blog written by Tom Pashby for the Green Careers Hub.
Image credits: 1- Shutterstock, 2 – AdobeStock, 3 – Shutterstock
