Page 21 - Connecting the Future How Connectivity and AI Unlock New Potential

Connecting the Future: How Connectivity and AI Unlock New Potential

closer to the end user to reduce latency and improve responsiveness. AI helps optimize last-mile
connectivity by:
• Predicting user behavior and pre-positioning data at local nodes using edge analytics, reducing
the need to route every request back to centralized servers.
• Dynamically allocating local resources including processing power and bandwidth based on
real-time demand, improving the efficiency of high-resolution content delivery. 55
• Proactively managing congestion and signal interference in last-mile access points (e.g., fiber-to-
the-home, 5G base stations, or community mesh networks), thereby ensuring consistent quality
of service (QoS) for applications like live AI video synthesis and interactive virtual experiences.
Moreover, AI helps balance backhaul and last-mile coordination, ensuring that the entire data
path from centralized cloud infrastructure to the user endpoint operates as a unified, intelligent
system. Advanced traffic classification algorithms and intent-based orchestration enable networks
to understand application needs (e.g., low-latency video rendering vs. bulk data uploads) and adjust
routing and caching policies accordingly.
This capability is not just a technical upgrade, it is an economic enabler. In developing regions,
last-mile AI connectivity can help leapfrog legacy infrastructure challenges by intelligently maxi-
mizing the performance of limited broadband or wireless deployments. For instance, AI-enhanced
last-mile solutions can:
• Extend the reach of broadband in underserved rural areas through optimized wireless mesh
networks and low-cost edge nodes. A study by the World Bank found that increasing mobile
broadband penetration by just 10% in developing regions can raise GDP growth by up to 1.4%,
a gain magnified when such growth is paired with AI-driven content and service delivery.
• Facilitate digital entrepreneurship by enabling creators in remote areas to participate in the
global AI economy, producing and selling high-resolution AI-generated content.
• Support e-government and digital health initiatives, where real-time video and secure data
exchange are essential for service delivery.
AI-managed end user networks are another critical component to last-mile connectivity. Advanced
Wi-Fi networks now allow users to use generative AI user interfaces to automate network spectrum
utilization as well as automatically identify and apply policies to every connected device on the
network. These systems also allow for the tracking of equipment in real time, the provision of
56
automatic updates to customers or employees, and seamless directions for end users to available
meeting rooms or desks.
In short, AI-driven last-mile connectivity is foundational to the next phase of digital economy growth.
It ensures that the benefits of generative AI and other advanced services are not limited to urban
centers or high-income users, but are accessible to diverse populations globally, unlocking new
economic opportunities and supporting accessible innovation.

2�2 AI-Driven Energy Efficiency

Despite major advancements in digital infrastructure, energy waste remains a significant challenge.
According to the International Energy Agency (IEA), improved efficiency measures could save up
to 3.5 gigatons of CO emissions annually – the equivalent of removing approximately 750,000 cars
2
from the road, based on average U.S. vehicle emissions. 57
While AI models demand computational power, they also offer tools to make digital networks more
energy-efficient. AI enables real-time monitoring and control of network components, allowing
systems to dynamically adjust energy usage based on demand. For example, AI can automate the
powering down of idle base stations during low-traffic periods or optimize bandwidth allocation
to reduce unnecessary processing loads. 58
A 2023 ITU report emphasizes that energy efficiency in AI-powered digital infrastructure is not only
a cost-saving imperative, but also supports sustainability; for instance, the report notes that ICT has
the potential to abate between 0.72 to 12,080 million tons of carbon dioxide equivalent (Mt C02e)
59
annually. Rather than simply contributing to rising energy use, AI can play a key role in enabling
more efficient network infrastructure. In this way, digital transformation and environmental goals
do not have to be in conflict. With the right tools and investments, they can strengthen each other.

15