The challenge of Artificial Intelligence for the mobile industry

It is reasonable to think that open source models are outperforming proprietary ones, I think no one can deny this, but it is also fair to think that, if it had not been for the huge multi-million dollar investments made by the US in Artificial Intelligence infrastructures, data centres, chips, graphics cards, etc., the new Chinese AI would not have been able to get as far as it has.

We must think beyond the one-off effects of a black Monday on the stock market, the alarm created and the sense of frustration of some leaders and understand it as an opportunity that both blocs and, ultimately, the rest of the world will be able to take advantage of. An opportunity that will allow us to maintain the balance, with its nuances, in this new space race, and thus, as DeepSeeK grows to billions of users and needs to incorporate new services such as video processing, it will need more processors and greater computing power in the future, and at the same time the big native American Artificial Intelligence technologies will be able to learn quickly from the Chinese open source code. I insist, we must understand this as an opportunity that will also allow us to have a more dynamic competition and a much more powerful and efficient technology; considerably optimise costs and consequently increase demand; balance the balance with respect to dependence on American hardware and make governments aware of the critical importance of continuing to invest in and promote the use of AI techniques in all sectors and at all levels, administrations, large corporations, as well as small and medium-sized enterprises, which in most countries, as is the case in Spain, account for more than 60% of the country’s jobs.

The mobile industry – operators, vendors, integrators, manufacturers – one of the industries that has invested the most in Artificial Intelligence from the beginning, must now take advantage of this moment, reinforce its commitment and try to efficiently optimise its investments in AI and make it profitable at the same time.

5G technology and AI are driving a significant increase in energy consumption globally, calling into question some of the sustainability promises made at the time of their conception. According to the GSMA’s latest report – Telco AI; State of the Market Q3 2024 – telecom networks and data centres each account for 1% of the world’s electricity, and cloud energy consumption is expected to increase by 30-60% by 2030. It is therefore necessary to continue research in this direction and find new AI solutions that, as mentioned above, allow us to optimise network and service operations while mitigating this high energy consumption. AI innovations such as RAN closures and dynamic spectrum management significantly optimise energy use and reduce operational costs.

Telco technology providers have a key role to play. They need to integrate AI as a fundamental part of their technology and operations to be competitive. Automation and continuous learning are critical to simplify and greatly facilitate the management of complex operations such as new deployments or predicting and automatically resolving incidents, while constantly learning from data, adapting to new service scenarios and efficiently optimising their technology.

Another major challenge for the mobile industry is to ensure security. It is necessary to develop robust, secure and resilient proprietary applications and AI systems capable of ensuring the privacy and protection of sensitive data, both in terms of network operations and data, and in terms of the subscribers themselves and their consumption patterns. International cooperation in this regard is crucial to share knowledge and experience, to develop common technical standards to ensure the security and interoperability of AI systems, and to have international regulation to mitigate the negative impacts of AI.

Last but not least, the challenge that mobile operators in particular face within the telecoms industry is to focus more on return on investment and revenue generation. These are undoubtedly the two biggest concerns for operators at the moment when it comes to using artificial intelligence. Reducing the cost of operating the network, creating autonomous networks that can adapt to demand at any time and make decisions in the event of an incident, automated customer service support, churn detection, etc., are some examples of AI-based solutions that can optimise costs and provide a gradual return on investment. However, revenue generation is a priority in the strategies of operators, who are focusing their efforts on finding new lines of business involving the monetisation of data on user behaviour and consumer trends, personalised marketing, etc.