Written by Sara Guo I 7 min
In a world teetering on the brink of ecological crisis, an unexpected hero emerges. Artificial Intelligence (AI) isn't just another player; it's rewriting the script for sustainability.
Why should you care? AI in sustainability applications could contribute up to $5.2 trillion USD to the global economy in 2030. And, can reduce worldwide greenhouse gas (GHG) emissions by 4% in 2030.
As we dive deeper into the AI-driven sustainability frontier, the possibilities are boundless. It's clear that AI is not just changing the way we approach sustainability; it's leading us toward a future where eco-challenges are met with data-driven precision and innovation, creating a more sustainable and harmonious world for generations to come.
Use Cases of AI in Sustainability
AI applications offer organizations precise insights into their environmental impact, the ability to forecast environmental challenges, and the means to minimize resource consumption while enhancing supply chain efficiency. This optimization extends to logistics, transportation routes, and inventory management, reducing emissions and waste. Let's now delve into companies that are at the forefront of AI-driven sustainability initiatives.
Example #1 — Tesla's AI in Electric Vehicles
Tesla's innovative use of AI extends far beyond the sleek design of its electric vehicles. They employ AI for adaptive cruise control, battery management, and the development of autonomous driving capabilities. This integration not only enhances the driving experience by offering safety and convenience but also significantly contributes to sustainability. Greenhouse Gas (GHG) emissions are perceived as the main contributor to climate change, so by optimizing energy usage and reducing emissions through AI-driven efficiency, Tesla is at the forefront of the electric vehicle revolution, paving the way for a cleaner and more sustainable transportation future.
Example #2 — Pachama's AI for Carbon Sequestration
Pachama, harnesses the power of AI and remote sensing to track carbon sequestration in forests and other natural ecosystems. Their cutting-edge technology, supported by machine learning algorithms, enables progress monitoring and early identification of potential risks. By analyzing regional characteristics, Pachama identifies ideal project areas that maximize the impact of climate investments on local communities, wildlife, and the environment. This AI-driven approach not only bolsters transparency and accountability in carbon offset projects but also accelerates our journey toward a more sustainable planet.
Example #3 — Xylem's AI for Water Treatment
Xylem, a global leader in water technology, leverages AI to revolutionize water treatment processes. Their AI-powered software predicts water quality and optimizes treatment processes in real-time. By doing so, Xylem not only ensures the delivery of safe and clean water to communities but also reduces energy consumption and operational costs.
Example #4 — Taranis' AI in Agriculture
In the realm of agriculture, Taranis stands out as a pioneer in AI applications. Utilizing advanced image analysis powered by AI, Taranis detects early signs of crop diseases, pests, and other stress factors in fields. This proactive approach allows farmers to take precise and timely actions, reducing the need for chemical interventions and optimizing crop yields. Taranis plays a vital role in sustainable farming practices, working toward UN Sustainable Development Goal (SDG) #2 Zero Hunger, while minimizing the environmental footprint of agriculture.
There is a growing abundance of Canadian AI startups focused on the agriculture space, demonstrating an opportunity for AI to alter how we approach our sustainability goals on a local level.
Ethical Challenges of AI in Sustainability
The biggest concern for many company executives right now is data security and privacy around using AI. However, when looking at sustainability, there are many stakeholders and complex considerations that must be accounted for.
- Cascading failures and external disruptions: Complex and nested AI systems risk experiencing shocks or “failures” that can have a ripple effect and disrupt our environment. For example, if an AI system responsible for managing energy grids experiences cascading failures or is disrupted by a cyberattack, it could lead to power outages or inefficiencies in energy distribution, which, in turn, could harm the environment and disrupt daily life.
- Non-anthropocentricity: AI should be used in a way that is not solely focused on human interests, but also considers the interests of other species and the environment. For example, fixating on enhancing agricultural practises purely through technologies like drones may unintentionally disrupt the delicate balance of life in the area.
- Environmental impact: The use of AI can have negative environmental impacts, such as contributing to electronic waste and the overuse of pesticides and fertilizers in agriculture. And, training AI models could require significant energy consumption, often derived from fossil fuels, thereby amplifying greenhouse gas emissions.
Unanswered Questions That Cannot Be Ignored
Evaluating AI's Effectiveness
How can we accurately assess the effectiveness of AI-powered sustainability initiatives and distinguish between genuine impact and greenwashing?
Assessing AI-powered sustainability initiatives demands ruthless scrutiny. The key lies in relentless transparency, robust measurement, and an uncompromising commitment to results. A trustworthy audit system, enforced by independent third parties, should be the gold standard. The metric should be the tangible, measurable change in the environment and society, not empty claims or superficial green labels. It's time to hold companies accountable for their sustainability promises with the same vigour as their financial performance. Anything less perpetuates the facade of sustainability while our planet's health hangs in the balance.
Energy and Resource Requirements of AI
What are the energy and resource demands of training and operating AI models at scale, and how can we make AI systems more energy-efficient?
According to a recent study conducted by MIT, the carbon emissions resulting from a single training session of a large language model, just one step in AI system development, amount to approximately 284 tons of CO2. To put this in perspective, it's equivalent to five times the carbon footprint of an entire car's life cycle, encompassing fuel consumption, and approximately 57 times the annual CO2 emissions of an average individual.
To mitigate AI’s resource demands, we need a seismic shift in AI model design and training methods. It's time to prioritize efficiency over extravagance. This means crafting leaner, more purposeful models, reevaluating the necessity of massive data sets, and optimizing algorithms for minimal computational waste. This is called the “Small Data” approach, looking at smaller training datasets, to reduce the energy consumption and resource-intensiveness of AI systems.
Job Displacement
What strategies are needed to ensure that AI advancements in sustainability do not lead to job displacement in certain industries or regions?
The job market is changing, and AI proficiency is a must. In the future, you are not competing with an AI robot; you are competing with someone who knows how to use AI.
Global Collaboration
How can nations, organizations, and researchers collaborate on AI-driven sustainability solutions while respecting intellectual property and geopolitical considerations?
In fostering global cooperation for AI-driven sustainability solutions, we must shift from narrow proprietary interests to collective well-being. Shared open-source platforms for AI algorithms and data can facilitate collaboration while respecting intellectual property. Establishing international consortia with transparent governance can provide a framework for knowledge exchange and equitable resource distribution. Crucially, geopolitical tensions must be set aside, recognizing that the environmental crises we face transcend borders. A united global effort, driven by a sense of shared responsibility, is the only way to harness the full potential of AI for sustainability in our interconnected world.
How to Proactively Champion Sustainability Within Your Organization
The easy answer to promote green practices is to keep up with emerging AI technologies in the sustainability space. But, there are more effective and high-impact solutions to enable an entire organization to start thinking about AI possibilities in sustainability.
But first, you need to understand AI and its potential.
Once you and your team feel comfortable conversing about AI, you want to take action.
What could that look like?
Hackathons. We believe that the best way to learn is by building - and that’s exactly what a hackathon can do for you and your business.
We have a strong track record of organizing 1000+ people global hackathons for Fortune 500 and leading AI companies. Some of our past clients include McDonald’s, Bank of Montreal (BMO), and Capgemini, for whom we recently organized a two-week generative AI hackathon, “GenAI Hackathon 2023,” to upskill their global employees in generative AI while generating 52 solid AI solutions with the potential to positively impact the company’s future decisions in this digital world.
In our hackathons, sustainability is consistently prioritized within the challenge streams. This commitment is vividly exemplified in our BMO hackathon and HSBC's five-week innovation sprint for new graduates, named 'Ideation Voyage.' We challenged participants to transcend conventional sustainability solutions and tailor their submissions to the unique parameters set by each company.
We believe that every company must proactively champion sustainability. It's imperative to acknowledge that the definition of a 'good' sustainability solution varies significantly between companies, contingent upon their distinct financial and non-financial constraints. Embracing this diversity is the bedrock of genuine sustainability leadership.
Open innovation works, and diversity lies at its core. Time-bound and high-competitive environments have proven to be successful as an agricultural advice AI won Canada’s recent Public Service Data Challenge that looked to challenge public servants across Canada to reimagine how to use data in an innovative manner.
Looking Ahead
AI is still a new frontier, and many company executives hesitate to leverage AI in their sustainability efforts.
What if we don’t do anything?
It will cost the US economy over $500 billion per year in 2090 for inaction.
In Harvard Business Review, former Unilever CEO Paul Polman reminds us, "Sometimes it's good to set unachievable goals because even if you fall short of it, you're 10x better than where you were before". Furthermore, AI development in this space is already seeing support from the Canadian government as the Minister of Agriculture and Agri-Food recently gave out $419,000 to an AI company to support the digitization of farming in the Canadian agriculture space.
In this era of AI and sustainability, embracing innovation is not just an option; it's a necessity. The potential rewards in economic savings, environmental impact, and societal progress are simply too immense to ignore. The time to act is now.
