The future of computing needs to be more sustainable with artificial intelligence (AI) playing a role in achieving this, even as the growing adoption of the technology fuels energy consumption.
Digital technologies such as AI can help identify ways to reduce emissions, such as optimizing power grids and developing sustainable supply chains, said Singapore's Deputy Prime Minister Heng Swee Keat.
AI models can analyze complex environmental data, find areas to improve, and drive more efficient, data-driven decision-making, said Heng, at the Alibaba-NTU Global E-Sustainability Corplab launch in Singapore.
Also: AI arm of Sony Research to help develop large language model with AI Singapore
Digitalization itself, though, can also significantly widen our carbon footprint, he said, noting that the tech industry alone currently contributes an estimated 1.5% to 4% of global greenhouse gas emissions.
"The digital revolution and green revolutions are intertwined," he added. "Just as the future of sustainability will be AI-driven, the future of computing must also be greener."
With energy consumption bound for growth as the use of AI climbs, everyone must use AI well, he said.
Also: How your inefficient data center hampers sustainability – and AI adoption
To do so, various measures are needed to ensure AI is deployed optimally while achieving sustainability, including policies, Heng said. Singapore's green data center roadmap, for instance, was introduced earlier this year to optimize its green energy use, efficiency, and computing capacity. It outlines the need for data center operators to work with enterprises to improve the energy efficiency of hardware and software, while energy suppliers have to scale up the use of green energy.
He added that the new corporate lab established by Alibaba and Nanyang Technological University (NTU) plays a role in boosting Singapore's research, innovation, and enterprise capabilities, particularly, in translating research insights into tangible real-world applications. Such collaboration will strengthen niche capabilities and facilitate interdisciplinary research, he said.
Today, Singapore houses more than 20 corporate labs across universities, Heng said, such as the ExxonMobil-NTU-A*Star corporate lab, whose efforts include developing efficient carbon capture and carbonization technologies.
Also: Microsoft Copilot to be integrated into Singapore's legal technology platform
The focus, for the Alibaba-NTU research facility is building sustainable digital applications, such as green AI models, to uncover new ways to cut energy use and minimize environmental impact, he stated. These can further support smart digital technologies and greater urban sustainability, he added.
This may be increasingly critical as organizations are unlikely to hold back their AI adoption even amid concerns about its impact on the environment.
Some 64% have expressed worries about how AI and machine learning initiatives will impact their energy use and carbon footprint, according to a study released by AI data platform WEKA and S&P Global Market Intelligence. Another 25% say they are very concerned about this impact.
Also: IBM will train you in AI fundamentals for free, and give you a skill credential – in 10 hours
Conducted in the second quarter of 2024, the survey polled 1,519 AI and machine learning decision-makers across enterprises, research organizations, and AI tech vendors.
Some 42% of respondents say their organizations have invested in energy-efficient IT hardware to address the potential environmental impact of their AI projects over the past year. Among them, 56% say this has had a high or very high impact, the study found.
Despite their concerns, 33% of respondents have AI projects that are widely implemented and driving significant value, compared to 28% last year. Respondents in North America lead the pack, where 48% have AI that is widely implemented, followed by Asia-Pacific at 26% and EMEA at 25%.
Also: Businesses still ready to invest in Gen AI, with risk management a top priority
In addition, 88% are actively investigating Gen AI outpacing other AI applications. For instance, 61% are exploring prediction models, while 51% are looking at classification, 30% are at expert systems, and 30% are investigating robotics.
Some 24% of respondents see Gen AI as an integrated capability deployed across their organization. Another 37% have Gen AI in production but not yet scaled, while 11% have yet to invest in Gen AI.
Also: The data suggests gen AI boosts software productivity – for these developers
The organizations, on average, have 10 AI projects in the pilot while 16 are in limited deployment. Just six AI initiatives are deployed at scale.
Asked about their biggest technology barriers to deployments, 35% point to storage and data management, while 26% cite computing, and 23% see security as an inhibitor.