In the realm of technology, artificial intelligence (AI) stands out amidst the diverse landscape, driving the current wave of excitement with the promise of significant productivity enhancements facilitated by cloud computing, expansive language models, user interfaces based on natural language processing, and the ability to generate text and images. Telecommunications providers are also embracing this trend, pivoting their messaging to emphasize AI for internal streamlining and customer-centric applications that could lead to new revenue streams. Notably, while the advent of 6G is still on the horizon in terms of standardization, initial discussions about the next iteration of cellular networks have centered around the concept of being “AI-native.” But what exactly does that entail?
Unpacking the Notion of “AI-Native” in Telecommunications
In a recent report titled “The AI-native telco: Radical transformation to thrive in turbulent times,” McKinsey and Company delved into the essence of this concept. To gain further insights into the vision and the practical steps required to actualize it, RCR Wireless News engaged in a conversation with Tomás Lajous, a Senior Partner at McKinsey and Company and one of the report’s co-authors. Lajous underscored the essential prerequisite of being cloud-native to embody the ethos of being AI-native, highlighting that telecom operators are yet to fully embrace cloud-native principles.
Lajous articulated, “The concept of a cloud-native telco is inherently linked to the notion of an AI-native telco. If we were to establish a telecommunications company from scratch today, the optimal approach would involve integrating AI at its core. This entails leveraging AI to inform virtually every decision-making process and operational framework, fostering a culture that wholeheartedly embraces AI across various functions, ranging from marketing and customer service to network management.”
Achieving this vision necessitates a robust technical infrastructure aligned with these objectives, with cloud integration emerging as a pivotal enabler for the AI-centric paradigm. Lajous emphasized the imperative of a cloud foundation to support the AI-driven facets of a telco’s operations.
Reflecting on the hypothetical scenario of launching a telecom enterprise “from scratch,” Lajous acknowledged the rarity of such opportunities, considering that even contemporary cloud-native initiatives like those spearheaded by Dish in the U.S. and Rakuten Mobile in Japan have not uniformly translated into commercial success. The journey towards becoming AI-native, he clarified, is not a predefined future state but rather a guiding principle for strategic deliberations.
He proceeded to outline the foundational elements that operators must prioritize, encompassing the modernization of network technologies across all domains, the transition to a 5G Standalone core, and enhancements to Operational Support Systems (OSS) and Business Support Systems (BSS). The latter aspect pertains to the effective provisioning, billing, and consumption of services within a cloud-native framework. Regarding the role of AI in catalyzing data monetization, Lajous acknowledged the existing challenges posed by regulatory constraints, user consent requirements, and other variables hindering operators from fully leveraging the wealth of personalized, contextual data at their disposal.
Leveraging AI for Enhanced Product Development and Customer Experience
Lajous underscored, “The significance lies not only in the conventional applications of profiling and marketing…Fundamentally, it revolves around enhancing the product itself. Historically, telecom operators have grappled with evaluating customer experiences, particularly within mobile networks…AI now empowers us to comprehensively monitor network activities and assess whether individualized needs are being met.”
Armed with this actionable intelligence, operators can refine their offerings, elevate the overall customer experience, and subsequently introduce distinctive value propositions. An additional hurdle pertains to the accessibility and coherence of data within operator ecosystems; for data to effectively fuel AI algorithms, it must be harmonized, structured, and readily accessible. Similar to many enterprises, operators encounter the challenge of disparate data repositories siloed across various functions.
“In the telecommunications domain, we have been ensnared in a detrimental cycle where inaccurate data leads to subpar AI outcomes, resulting in diminished emphasis on data generation, thereby perpetuating the cycle of inadequate data quality,” Lajous remarked. “However, we are breaking free from this cycle through two pivotal approaches. Firstly, leveraging digital twins as the cornerstone of AI strategies instead of relying on use case-specific datasets…Secondly, the emergence of gen AI…Gen AI effectively addresses this predicament.” He illustrated this point with an example of circuit inventories and associated voluminous documentation. “Gen AI empowers us to amalgamate this wealth of information into a cohesive entity, akin to a data product or digital twin, which can then undergo AI modeling for circuit data analysis. This exemplifies how, through digital twins and gen AI, we can bridge the existing gap.”
