AI at the Edge: Unlocking AI Efficiencies and User Experience with AI-Enabled Personal Devices?
Every AI interaction processed in the cloud costs tokens. Across a workforce scaling AI adoption, those costs accumulate fast – and for many organisations, a significant share of them are being generated by tasks that did not need to go to the cloud at all. Modern AI PCs and mobile devices carry GPU, NPU, and TPU capabilities that are largely sitting idle. That is compute your organisation has already paid for.
This track explores how to put that hardware to work as part of your AI cost strategy. By shifting targeted, high-frequency AI tasks onto the processing power already embedded in end-user devices, organisations can drive material reductions in token spend while improving the speed and responsiveness of AI interactions for end users. The approach applies across the full device estate – from AI PCs handling knowledge work and developer tooling to mobile devices supporting field and frontline workers. If those devices are still being treated as endpoints rather than as part of the AI compute layer, this session will reframe that conversation.