The AI supercycle offers a transformative growth opportunity for telecoms providers. To seize it, networks need to provide deterministic connectivity. A critical part of this shift will be investing in highly adaptive autonomous networks that can anticipate changing traffic mixes and operate at machine speed to enable rapid service creation and seamless execution, writes Vivek Jaiswal, the senior vice president of Autnomous Networks at Nokia.
Nokia’s AI-native automation framework was built for the AI era, combining data, orchestration and resource automation to deliver essential and trusted networks in the age of AI. Importantly, AI isn’t just more traffic, it’s different traffic and the AI supercycle is unfolding in waves.
Generative AI is already mainstream. Agentic AI is making it possible for goal-oriented software agents with ‘human-like’ reasoning capabilities to drive business and operational outcomes, enabling them to interact and coordinate with each other. Physical AI will make latency and reliability non-negotiable, merging the digital, physical and human worlds.
Nokia’s global wide area network (WAN) forecast projects AI traffic will reach 921 exabytes per month by 2034, growing at an annual rate of 23% and accounting for around 30% of total global WAN traffic.
But the bigger disruption is qualitative. AI workloads are throughput-hungry, latency-sensitive and, crucially, bursty and distributed across datacentres, metro networks, edge clouds and enterprise sites.
Vivek Jaiswal