As they move forward with the adoption of generative artificial intelligence (GenAI), telcos need to explore use cases that actually add value. Otherwise, you may have to deal with future problems.
Peter Jarich, head of GSMA Intelligence, pointed out that AI is not a new technology for telcos, adding that GenAI has democratized the use of such tools and sparked widespread interest in the tool. said that it has become a lot of.
Also: 5.5G is touted as the network that will deliver improved enterprise connectivity
The key question now is how telcos should take advantage of new opportunities. He pointed to 5G, which is the fastest growing mobile technology but has not been as successful in increasing profits and revenues for carriers. This is why so much discussion continues to revolve around monetizing 5G and leveraging its capabilities to provide useful services.
He further pointed out that interest in GenAI among carriers remains low, with 56% still in the testing phase and the number of commercial deployments low.
There are good reasons for carriers to move slowly, however, as they operate networks that support critical infrastructure and need to consider the regulatory implications if downtime occurs. The challenge now is to consider how to move beyond the testing phase of GenAI. Preparations are underway for commercial launch, he said.
There’s also so much hype surrounding GenAI that carriers need to cut through the noise and think about how they can leverage this technology to deliver real value, said a conference this week. Jarich said speaking to ZDNET ahead of Mobile World Congress (MWC). Barcelona, Spain.
Related article: Best gadgets seen at MWC 2024: smart rings, concept phones and transparent laptops
The biggest obstacles facing carriers are uncertain return on investment (ROI) and technology maturity. He urged the industry to identify his GenAI proof of concept with revenue-generating potential and establish one or two compelling use cases.
Multiple players are currently vying for market share, pushing a wide range of products including different AI chips and capabilities. Tying GenAI as a marketing catchphrase to everything could lead to consumer disillusionment if these services fail to provide real value, he said.
Therefore, the industry needs to be careful about how it wants to market GenAI or risk customers losing trust in the technology, Jarich said. He emphasized the need for clear messaging and a basic understanding of the tools available.
With a focus on GenAI, broader AI should focus on reducing operational costs and providing better customer support. He noted that carriers want to use these tools to come up with new services that can increase value and build stronger connections with consumers.
He added that there was also a movement among device manufacturers such as Samsung and Google to offer AI-equipped devices. With smartphone sales plateauing or declining, these market players have had to consider ways to encourage consumers to buy new models. This led to the introduction of foldables, for example, he said.
Related article: AI’s latest tricks: Better KPIs to measure business success
They now expect similar things from GenAI, such as improving search and adding more useful features, he added.
This drives the need for open APIs to allow developers to build GenAI tools that take advantage of 5G features such as low latency, further driving new use cases for 5G and driving demand and traffic. says he.
GSMA predicts that 5G connections will increase from 1.6 billion to 2.1 billion by the end of 2024.
Jarich noted that use cases can vary by region and market, giving local carriers further opportunities to find new revenue. And since most GenAI services run on the cloud, operators can play a differentiated role in facilitating them. For example, we can provide support for edge computing. This will be important for some of his GenAI services, such as real-time language translation and small-scale large language models (LLMs) that can be hosted on devices.
Dedicated platform model for carriers
Domain-specific LLMs are also available to identify GenAI use cases specific to the needs of specific sectors, such as finance or healthcare.
Related article: 3 innovative ways edge computing and 5G are transforming healthcare
Huawei this week announced a communications infrastructure model that carriers can use to improve operational efficiency and optimize network productivity. The company’s telecom infrastructure model covers two major applications: role-based co-pilot and scenario-based call agent, said Yang Chaobin, Huawei board member and president of ICT products and solutions at MWC. He said this beside.
AI models can enhance natural language interactions, analyze complex processes, and adjust operations to provide a better customer experience for different roles and scenarios, Yang said. These are customized for roles such as network optimization agents, user experience agents, and fault management.
He added that this foundational model can also power autonomous networks and provide carriers with three core capabilities, including service provisioning and network maintenance.
Telcos believe that by combining autonomous networks with technologies such as AI, big data, cloud, and edge computing, they can deliver services faster and at lower cost, making them easier to deploy and manage. For example, the TM Forum has defined fully automated networks for vertical industries, with an Autonomous Networks project aimed at enabling communications networks that are “self-configuring, self-healing, self-optimizing, and self-evolving.” It is operated.
Related article: How Renaissance technologists are connecting the dots between AI and business
Yang emphasized the importance of automation, noting that carriers are facing an increase in Opex (operating expenses), which has risen to about 70% as a percentage of revenue. Software-defined, autonomous networks could help alleviate these cost pressures, he said.
He added that Huawei’s telecom infrastructure model can enhance service provisioning use cases where administrators can access “accurate multimodal assessment and rapid service provisioning.” AI model optimization capabilities will further facilitate user experience assurance use cases, while cross-process analysis and dialog-based support will enhance troubleshooting cases, he said.
As AI applications leverage more data, data security will become a bigger focus, allowing operators to act as custodians of user data, Jaric said. He also noted that as more GenAI-generated content and services are adopted, companies need to ensure their networks can handle the increased traffic.
He also said that because telecom companies already have vast amounts of data about their customers, they can leverage AI to manage bespoke services and better provide services that meet users’ specific needs. That’s the lesson they can take from the 5G experience, he added.
Also, what is 5G Home Internet? Here’s what you need to know before signing up
Bruno Zhang, Huawei’s cloud chief technology officer, said that by driving “a new level of intelligence” with foundational models and GenAI, enterprises can bring AI-generated content to production environments and software engineering. He said he will increasingly use it for support. But building your own foundational model will be difficult because it requires “systematic innovation,” Zhang said at the Chinese Vendor Cloud Summit held alongside MWC.
He said Huawei wants to help by providing AI-based models to power applications and its own cloud services. The Chinese tech giant also aims to ease the adoption of AI by providing key components such as AI-native storage and AI-powered data.
He said Huawei’s Pangu LLM, released last year, includes industry-specific models trained using industry data to provide industry-specific scenarios and tasks, such as autonomous driving and weather forecasting. It pointed out.
Singapore-based Irene Yu reported to ZDNET from the 2024 Mobile World Congress in Barcelona, Spain at the invitation of Huawei Technologies.
