Is machine learning sustainable?

Global agenda 21 March 2024

Are AI and machine learning sustainable technologies? Esade Professor Irene Unceta hosted a panel of experts at 4YFN to explore the economic, social and environmental perspectives of this crucial question.

Do Better Team

The pursuit of sustainable development offers many opportunities – but they don’t come without accompanying challenges. During a discussion at the 4YFN conference, Esade Professor Irene Unceta posed the crucial question: are AI and machine learning sustainable technologies

Expert panel Idoia Salazar (founder and president of the Observatory of the Social and Ethical Impact of Artificial Intelligence), Paloma de la Puente (partner and sustainability expert at Conese) and Joan Ramon Mallart (director of IBM Consulting Experiences), each brought a unique perspective to the pursuit of sustainable development and the issues presented by the rapid rise of AI

Accelerating solutions

As a specialist in the design of user experience-utilizing technologies including AI, IBM’s Joan Ramon Mallart brought an economic perspective to the ongoing debate. 

“What we are doing at IBM Consulting is creating and co-creating,” he explained. “We work with knowledge to accelerate the ideation process and solve the large, complex problems our customers face.” 

AI can influence corporate strategies to introduce innovative sustainability perspectives to production

Ultra-personalized communications, media and technology and even scouting for future sports stars are all real-world examples cited by Mallart. “We use the power of data to reveal hidden patterns that help us, for example, to identify future talent, like the next Messi when he or she is 12 years old,” he said. 

Hybrid cloud solutions and open-source models will create an ecosystem that will benefit “all levels of companies,” Mallart said.  

Introducing efficiencies

Paloma de la Puente, who has over 15 years of experience in technology applied to sustainable development, advises companies on how to integrate sustainability strategies into their business models and reduce their environmental impact.  

She sees “a very big chance” for AI to influence and shape corporate strategies for sustainability by looking beyond the targets of compliance and risk management and introducing innovative sustainability perspectives to production.  

De la Puente described two sectors she’s currently advising on the deployment of sustainable practices in Spain: cosmetics and auto parts. 

AI can be a tool to address challenges like climate change, health, and poverty

“Both of them are very involved in technology,” she explained. “But in environmental, social and governance issues there are opportunities to look at their emissions, their energy, their waste management. 

“We’re working with them to integrate AI to reduce emissions and control the levels of component manufacturing. We don’t need to produce more than is necessary. AI and machine learning introduce a lot of benefits and efficiencies.” 

The social dimension

Idoia Salazar is a leading expert in ethics and legislation in AI. Actively involved in shaping policy at both national and international levels, Salazar said AI should be treated as a complement to human knowledge — not a substitute. 

“We have a lot of challenges in front of us, including climate change, and I think artificial intelligence is the tool needed to address these challenges,” she said. “But we have to start from a very objective point of view. 

“We need to see AI as a tool to complement things that we as humans cannot address. Health, climate change, poverty; there are many ways we can use AI to build solutions to very specific problems. 

“In the health sector, for example, it can help us to personalize treatments for diseases like cancer. But it can also help us to treat the rare diseases we don’t have enough money to fund.” 

Education is key

However, Salazar stressed, there is a very strong need for education – a point that host Unceta put to the panel’s tech expert, IBM’s Mallart. 

“To try and mitigate the potential negative effects, we need to know how to use it as users,” she told him. “We need to know how to interact with it as developers. We need to know how to design it. And as businesses, we need to know how to integrate it.” 

There is a very strong need for education to mitigate the potential negative effects of AI

Mallart provided assurance that integrating technology into an educated society is an obligation the tech giants take very seriously. Sustainability is a team sport, he said: open access products provide tools all companies can use to develop effective ethics strategies. 

“We strongly believe that really there is no dilemma for companies anymore between profitability and sustainability,” he stressed. 

Predicting the future

Ultimately, the panel agreed, historical data and the information that continues to be amassed exponentially can be utilized by AI and machine learning to identify solutions to both global and local issues

“The number one starting point is always data,” said Mallart. “With complex problems like climate change, we can provide code to understand exactly what's going on and make accurate predictions.  

“In the past, we only had 24 hours before knowing a hurricane would arrive at a specific location. Now, we can see the impact we have on deforestation, floods and wildfires. We can understand and tackle the motivations of these challenges and – hopefully — leverage our data to reverse them.” 

Personalizing everything means to lose sight of some of our own privacy

The use of natural resources and the emissions from homes, schools and communities can also all be made more efficient by the use of AI, according to sustainability expert Paloma de la Puente. And, she added, the efficiencies far outweigh the output of the technology used to manage them. 

Personalization v privacy

While acknowledging the data-driven benefits, ethics expert Idoia Salazar also was keen to remind the panel of ongoing privacy concerns

“If we are going to personalize everything that in some way is going to be more comfortable for us, of course that means that we have to lose sight of some of our own privacy,” she said. 

“This is very easy to manipulate, as we’ve seen when we sit in front of the television and are given recommendations of what to watch, or in political campaigns. I think that is the real problem we have to face with AI.” 

All written content is licensed under a Creative Commons Attribution 4.0 International license.