After several days of recovery and retrospection, the following are my key impressions regarding MWC 24.

·     While AI was a key theme, the near-term practical use of #GenAI or #LLMs to help mobile operators, vendors and others in the industry seems limited to customer support, customer experience (CX) and helping development teams with coding. The potential for LLMs to help with network planning and optimization and with security and fraud prevention is strong. But given the amount of data involved, the effort to getting the data usable for LLMs plus the massive amount of computing required for training and running the models, it seems like a very expensive multi-year or even generational effort. Network operators and vendors are saying all of the right things in terms using and adopting GenAI, but significant barriers remain in how much it can help operators’ topline and profitability in the near term.
·     
·     The conference opened with news of the progress of GSMA’s Open Gateway Initiative for a second consecutive year. Depending on your perspective, this could either point to the importance of the effort to reenergize operator business models to drive top line growth or that there is no other significant news, reflecting the current stasis of the mobile market. Nevertheless, we believe that the Open Gateway Initiative is important in changing operator business models as well as enabling enterprise communications and IoT connectivity platforms that can be used by enterprises and developers to create new applications and services. But it will be a gradual, multi-year transition.

·     Outside of China and a few operators in North America and India, progress on 5G SA core remains slow, yet the functionality that 5G SA core enables, remains critical to help drive new operator business models, enable automation, and modernize current services. We expect strong growth in 5G SA network functions in 2024 albeit coming from a relatively small base in 2023.

·     I had several discussions about whether splitting the 5G core network into 5G NSA (which leverages the 4G EPC) and an entirely new 5G standalone core network (5G SA), was in retrospect, the right thing to do. Generally, vendors lamented that the option of 5G NSA set operators further behind hyperscalers in terms of automation, costs, and scale, while operators appreciated that they were able to more economically deploy 5G NR and that the technology (“cloud-native”) underpinning 5G SA was/is not ready technically, or economically, for telecom workloads. Most people, and I agree, do not think we will be seeing “6G SA” and “6G NSA.”