Session
Organizer 1: Technical Community, Asia-Pacific Group
Speaker 1: Sumon Ahmed Sabir, Technical Community, Asia-Pacific Group
Speaker 2: Edmon Chung, Technical Community, Western European and Others Group (WEOG)
Speaker 3: Sabhanaz Rashid Diya, Civil Society, Asia-Pacific Group
Speaker 2: Edmon Chung, Technical Community, Western European and Others Group (WEOG)
Speaker 3: Sabhanaz Rashid Diya, Civil Society, Asia-Pacific Group
Format
Roundtable
Duration (minutes): 90
Format description: Conceptualised as a brainstorming activity with multiple groups and stakeholders, this session is proposed to be held as an ongoing conversation between stakeholders involved. There will also be differing perspectives on the topic at hand, which will need to be discussed in depth as an open dialogue. Hence, we request a 90 minute session.
Duration (minutes): 90
Format description: Conceptualised as a brainstorming activity with multiple groups and stakeholders, this session is proposed to be held as an ongoing conversation between stakeholders involved. There will also be differing perspectives on the topic at hand, which will need to be discussed in depth as an open dialogue. Hence, we request a 90 minute session.
Policy Question(s)
[1] What role can coding efficiency in software, applications, AI models, and cloud computing play in addressing sustainability and environmental concerns?
[2] What role can the software development industry play?
[3] What steps can be taken from the user side (companies, individuals, etc.)?
[4] How can policymakers and governments incentivise efficient and sustainable coding practices?
What will participants gain from attending this session? This workshop will provide an opportunity to everyone present to understand the role of efficient coding in making software, applications, AI, and cloud computing more sustainable. It will also provide a chance for various stakeholder groups to come together to understand their respective expertise and contributions to this endeavour. A collective shift in mindset will require tech and energy industry leaders, the technical community, policymakers, and the international community to brainstorm complementary solutions to adopt computing practices.
SDGs
Description:
As digital development efforts scale up across the world, we must ensure sustainability efforts do not widen existing digital divides and environmental damage. Artificial intelligence (AI) and cloud computing are power intensive, and research has shown that training an AI model emits significant carbon. Even a single AI search emits about four to five times the amount of carbon as an input on a search engine. As the world experiences rapid AI and cloud computing developments, emissions stand to skyrocket without necessary, timely, strategic, and intentional interventions. Decades ago, programmers and coders worked within severe hardware limitations, and were compelled to write lean, optimised code. However, with increased processing and connectivity capabilities, we have embraced bloatware and inefficient coding in software, applications, and systems. This inefficiency and unwillingness to be sustainable in the way we code has led us directly to more power-intensive applications, more server load, and greater environmental impacts. When thinking about the environmental and sustainability impacts of the Internet and digital development, we often think of hardware efficiency and sources of energy being used. However, software efficiency is an equally critical piece of the puzzle. AI can be part of the solution. Innovations in low-power AI inference and models, as we have recently seen with models such as DeepSeek, show that we can innovate with lesser when we have to. This session will be a collective brainstorm with tech and software companies, R&D firms, and other businesses to firm up our collective shift in mindset – coding efficiency should be a sustainability goal, just as using renewable sources of energy in data centres. We will also engage policymakers to discuss strategic shifts in policy as well as cross-border collaboration to adopt better computing practices.
As digital development efforts scale up across the world, we must ensure sustainability efforts do not widen existing digital divides and environmental damage. Artificial intelligence (AI) and cloud computing are power intensive, and research has shown that training an AI model emits significant carbon. Even a single AI search emits about four to five times the amount of carbon as an input on a search engine. As the world experiences rapid AI and cloud computing developments, emissions stand to skyrocket without necessary, timely, strategic, and intentional interventions. Decades ago, programmers and coders worked within severe hardware limitations, and were compelled to write lean, optimised code. However, with increased processing and connectivity capabilities, we have embraced bloatware and inefficient coding in software, applications, and systems. This inefficiency and unwillingness to be sustainable in the way we code has led us directly to more power-intensive applications, more server load, and greater environmental impacts. When thinking about the environmental and sustainability impacts of the Internet and digital development, we often think of hardware efficiency and sources of energy being used. However, software efficiency is an equally critical piece of the puzzle. AI can be part of the solution. Innovations in low-power AI inference and models, as we have recently seen with models such as DeepSeek, show that we can innovate with lesser when we have to. This session will be a collective brainstorm with tech and software companies, R&D firms, and other businesses to firm up our collective shift in mindset – coding efficiency should be a sustainability goal, just as using renewable sources of energy in data centres. We will also engage policymakers to discuss strategic shifts in policy as well as cross-border collaboration to adopt better computing practices.
Expected Outcomes
[1] Multistakeholder dialogue on the need to address inefficiency in coding practices and bloatware to further sustainability
[2] Identifying the role of various stakeholder groups in addressing the issue
[3] Brainstorming complementary solutions to be implemented by various groups to meet the goals identified
[4] Potential partnerships identified to interrogate coding practices and identify inefficiencies
Hybrid Format: This session is planning to have both onsite and offsite moderators, onsite and offsite speakers, as well as audience that will join in in-person and online. The onsite and offsite moderators will work with one another to ensure a seamless stream of dialogue and knowledge sharing between both speakers and audiences, regardless of how they are joining the session. We will ensure that online speakers and audiences have the tools and space to interject, speak to their ideas, and ask questions of the group onsite and vice versa.