Recently, on Environmental Variables, Anne Currie, a Green Software Champion, and Asim Hussain, Executive Director of the GSF, discussed the intersection of AI, sustainability, and the changing world of green software legislation.
Let's look at the core components of the conversation and what they mean for the future of sustainable software.
Sustainability Trumps Cost, but…
Sustainability is now perceived as more than just an environmental aim; it's a crucial measure of business success. According to the recent Green Ops Survey results, for most IT professionals, sustainability emerged as a higher priority than cost.
The survey signals a significant mindset shift in the industry towards minimizing emissions associated with running software. Respondents of the 2023 State of Green Software survey also chose security, reliability, and sustainability over cost, in that order, underscoring this trend.
While we're experiencing a shift in individual priorities toward sustainability, this often diverges from official expectations or perceived norms. We still need to find a formal way to recognize sustainability as a strategic advantage alongside security and resilience efforts, which could optimize costs and resources.
"If an organization ran a security review and turned off all its unused machines, they would boost security and sustainability performance."
It's time we close the perception gap and align sustainability with security and resilience efforts to improve cost and resource optimization.
Cost as a Proxy for Carbon
While lowering costs may lead to reduced carbon emissions, it may not since software systems and energy grids are becoming more complex. Projects such as the Green Software Maturity Matrix are good indicators of the varying stages of green software adoption. They can help understand the potential cost benefits as green software projects mature.
Companies should use financial metrics to promote environmental sustainability in the tech sector and tailor their approach to consider the unique environmental impacts of both development and operations.
This approach can support teams in navigating scenarios where reducing prices may involve using more carbon-intensive energy sources while selecting regions with cleaner energy, which may entail higher costs.
Standards for Emissions Reduction
Scope one, scope two, and scope three are not immediately critical to most people, who simply need to ensure unused machines are powered off—a basic measure that helps right-size usage and reduce software's carbon emissions.
For hyperscalers and companies doing more in the space, there is a need for significant improvement toward clear and relevant emissions classification for software. While the Tech Carbon Standard shows promise for understanding technology emissions, there is still a need for standards and classifications–protocols tailored for software companies. The existing GHG Protocol, designed primarily for manufacturing entities, does not fully address the unique emission profiles of software operations.
Data Centres and Green Energy
If we can judge by recent history, the energy demands of data centers and AI will skyrocket in the next decade.
While the tech industry is now more conscious of AI’s high energy consumption, it’s worth asking: Even when AI uses green energy, should we remain concerned about consumption?
At the GSF, we're working towards a future where data centers operate synergistically with green energy sources in regions abundant in energy and low in demand, thereby alleviating resource competition. Tools like Impact Framework offer a means to measure, verify, and openly disclose software emissions and other environmental impacts like water usage and air quality.
We're asking organizations to use tools like the SCI Specification and Impact Framework to examine their software emissions more closely and improve our trajectory towards a net zero future.
You can hear the full episode on Environment Variables.
This article is licenced under Creative Commons (CC BY 4.0)