The environmental impact of ASSIST
Environmental impact, efficiency and the steps Wyser is taking to reduce the computing resources required by ASSIST.
AI services require computing resources. Processing audio and generating summaries uses computing power and electricity, which has an associated carbon footprint.
How are we reducing the environmental impact?
We continuously work to improve the efficiency of how ASSIST processes and handles data. We review our processing methods and look for ways to reduce unnecessary computing and data processing while maintaining the quality and performance of the service.
Looking into the future, a model designed specifically for the tasks ASSIST needs to perform could require less computing power per task than a larger general-purpose model. We are still exploring this approach, so we cannot currently quantify the potential reduction in environmental impact.
How is ASSIST becoming more efficient?
The way we run ASSIST is already significantly more efficient than it was 12 months ago. We expect this to continue as we improve our processes and learn more about how to reduce unnecessary processing.
Our wider approach
Wyser takes its environmental responsibilities seriously. We operate with remote working as the norm, which reduces the environmental costs associated with travel for our own operations.
We also continue to look at how we can use resources efficiently across the wider business.