Using AI-generated content to build an enterprise metaverse

The emergence of the enterprise metaverse is a huge opportunity for organizations seeking to create engaging content and services for a growing variety of business needs. Moreover, organizations that invest resources in laying its early foundations have the opportunity to build an architecture that could have a permanent impact on the nature and dynamics of the metaverse for many years to come. For the telco industry, the creation of an enterprise-level metaverse requires housing network services that can be easily accessed and manipulated. Tools for resource management, software and product development, and customer engagement could all be used to support customers in their mission to build and deploy advanced next-generation services.

This is the purpose behind the Revolutionizing telecom operations empowered in the metaverse by AI generated content - Phase III Catalyst - the third phase of a larger group of projects focused on helping people to transcend the physical world in commercial settings and workplaces via 3D digital environments. In the initial phase, the project established AI-supported avatars which could be used in the metaverse to assist with a variety of customer service issues, so that their human counterparts didn’t have to. The project’s second phase concentrated on creating a marketplace or ‘business hall’ for organizations to provide products and the means for avatars to engage with each other. Now, the project’s third iteration is providing a means for third parties to easily build and develop immersive, engaging content and services.

Doing so involves investigating how AI-generated content (AIGC) could be incorporated for metaverse applications, the ultimate aim being to create an industry workspace inside the metaverse. To support this, the project team is designing and implementing AI copilots based on generative pre-trained transformer (GPT) technologies. These copilots enable CSPs to optimize network infrastructure, streamline software development processes, and deliver exceptional customer service, ultimately enhancing creativity and productivity in metaverse environments.

Firstly, to improve and save cost on network operations, the project developed NetGPT. This AI model allows CSPs to optimize network infrastructure design and operation effectively within the metaverse by using AI-generated content to simulate and analyze network scenarios, identifying areas for cost optimization, resource allocation, and efficient network management. By using NetGPT in metaverse scenarios, CSPs have been able to make considerable cost savings and maximize network performance.

To support the creation and development of engaging metaverse software for the metaverse, the project team has employed a second AI model CodeGPT. CodeGPT is specifically designed to enhance software development productivity within the immersive workspace of the metaverse. It automates repetitive coding tasks, provides design recommendations, and assists in aligning requirements, design, and implementation processes. CodeGPT has thereby enabled CSPs to accelerate software development cycles, reduce manual input, ensure code consistency, and drive innovation. With the telecoms industry already looking ahead to 6G, this approach could be used to great effect in efficiently building virtual network models before physical deployment – comparable to how the automotive industry today creates innovative new car models in the metaverse before production commences.

To improve customer service and support revenue growth, the project team has created an additional AI model AutoGPT. The purpose of AutoGPT is to enable CSPs to deliver exceptional customer service within the metaverse. It incorporates natural language processing and conversational AI, enabling personalized recommendations, addressing customer queries, and streamlining customer service operations. By using AutoGPT, CSPs can enhance customer satisfaction, drive revenue growth, and improve customer retention.

The Catalyst has had tremendous success so far in its task of creating a metaverse workshop. By using AI to support content generation, those working inside have been able to create vivid constructions of real-world and metaverse-based products and are able to test and understand their physical properties with far greater speed and accuracy. All three AI models deployed have achieved their objectives: early analysis of the deployment of NetGPT shows it has resulted in a 20% time-saving on network planning, while software development has also been improved, with the project team reporting that software development time has decreased by 55%. AutoGPT has also been reported to achieve early success and has increased customer loyalty and satisfaction by 5%, while also contributing towards a double-digit revenue increase in metaverse content in the last year.

According to Nan Xu, Director of Empowerment Division of IT, Big Data and AI Center, Cloud & Network Operation Department at China Telecom, “by harnessing the potential of AIGC in the metaverse, telco operators can transform their operations, improve efficiency, and deliver superior customer experiences. This project paves the way for telcos to innovate, adapt to changing market dynamics, and secure a competitive edge in the telecom industry." With AIGC showing such promise at an early stage, the telecoms industry is ideally placed to play a key role in metaverse architecture and enabling the creation of rich, engaging content within it.