As Microsoft and Google jostle aggressively for leadership in a generative AI arms race that seems likely to surpass short-lived hype around other recent buzzwords (think metaverse and NFTs), unsurprisingly, Middle Eastern powerhouse e& was one of the first operators to pioneer the technology on behalf of the telecoms sector.
The company formerly known as Etisalat is not shy of making a big statement, particularly when you recall how it was first out of the gate (along with fellow Middle Eastern players) in claiming the launch of 5G in mid-2018, usurping rivals in Asia and North America which had long been tipped to lead the way.
While many saw that 5G launch purely as a marketing ploy, the company has since self-styled itself as a global technology group, moving away from its traditional operator roots.
And in late March, it announced it will integrate ChatGPT, the large language model backed by Microsoft through an investment in maker OpenAI, into its internal processes to “elevate customer experience”.
Perhaps most significantly, the company said the integration with ChatGPT marked an important step in the use of AI in the telecoms sector and the operator’s early leap into adopting the model raised an important question.
As technology companies now scramble to find a place in the generative AI race, what role will it play within the telecoms sector and, equally, what role will the sector play in its development?
Generative AI to generate revenue
Speaking to Mobile World Live (MWL) Nischal Khorana, VP global at Frost & Sullivan, explained while leading operators have been exploring AI to enhance customer experience for a while, generative has the potential to be a “game changer” in actually creating positive value for businesses overall.
Khorana said this can be done in a number of ways, for example using the technology to enable personalised conversations with product recommendations and offers to drive greater sales and reduce churn. Longer term, he also highlighted an opportunity in helping operators gain a competitive advantage for telecoms services, with providers able to strengthen their enterprise business and analytic service offerings.
“The disruptive impact across enterprises and the early stage of adoption can create potential growth opportunities across multiple ICT services including advisory, data monetisation, data management, deployments and governance.”
Khorana noted while the opportunities clearly exist for operators, rapid evolution of the ecosystem needs to be considered and actually generating revenue from services encouraging enterprises to adopt the technology “will require a significantly different approach backed by strategic partnerships”.
Kester Mann, director, consumer and connectivity at CCS Insight, agreed the initial operator use case will come from improving customer care, noting younger customers are likely be attracted to generative AI-powered services because they “are more open to new technologies and less wedded to traditional forms of communications”.
Mann also argued the technology could put further pressure on operator retail stores, potentially hastening closures, pointing to research from CCS Insight conducted in 2022 which found three of the top-four responses about planned operator interactions would come in the form of digital.
While being heavy adopters, Mann is not expecting operators to play a lead role in development, instead tipping them to seek partnerships with the likes of Google and Microsoft to deploy the technology across their operations.
Indeed, with the vast amount of investment being made by companies like Google and Microsoft to incorporate AI tools into their consumer offerings, it would naturally make sense for operators to jump on the bandwagon.
Volam Samrat, project manager at GlobalData Technology, pointed out a few areas where operators could actually play a significant role in development, albeit in a supporting capacity.
Firstly, he noted the access operators have to vast amounts of data on consumer behaviour and network performance, which could be used to train AI models with “accurate and reliable predictions, recommendations and insights”.
A second possibility lies in infrastructure support. Samrat noted operators could provide the necessary infrastructure, such as cloud computing platforms and data centres, to support development given the significant amount of computing power and storage capacity required.
And a third potential lies in security and privacy, with operators ensuring appropriate measures can be in place to process AI applications that involve processing sensitive data such as personal information.
Use cases for telecoms and the wider world aside, questions have been raised about the dangers of the rapid rise and adoption of generative AI systems.
And while a Middle Eastern heavyweight made it clear it saw the technology as playing a role in its services, it was the CEO of a European telecoms giant who felt compelled to add to a chorus of lobbying about the dangers of generative AI coming from leaders in tech, governments and regulators.
Telefonica CEO and GSMA chair Jose Maria Alvarez-Pallete issued a blog to warn of the existential risks associated with development of an out-of-control generative AI model which could prove harmful to humans either biologically or socially.
He also warned companies that develop generative AI did so “without knowing how to stop the process” when the model requires a degree of restrained autonomy.
His comments followed an open letter backed by more than 1,000 signatories, including Twitter CEO Elon Musk no less, calling for a six-month pause in the development of AI models more powerful than ChatGPT successor GPT-4.
Mann believes calls for a moratorium on the subject will only grow louder, but the real challenge lies in how to enforce it.
“This could lead to a split between people and organisations who want to progress the technology and those who’d rather take a pause. Finding a consensus could prove elusive,” he said.
Khorana concurred, adding generative AI will only continue to grow and evolve to create greater value, but a collective focus by governments, technology companies and other stakeholders is expected to “gain greater emphasis and drive sustained adoption.”
The hype train
You can’t open a technology website today without being inundated with headlines warning about the impact of generative AI on society, the latest tools adopted by the likes of Microsoft to boost their offering, or indeed new platforms being launched by other companies to compete.
So the chatter about this topic is well and truly in full swing.
Mann notes telecoms, for one, has a “bad habit of over-hyping things, the metaverse being the most obvious recent example”, and “escalated levels of excitement is likely to give way to more healthy dose of reality before too long”.
Samrat, too expects the hype, like with any technology trend, to eventually subside, “as the industry and users become more familiar with its capabilities and limitations”.
However, the pair, along with Khorana are pretty confident generative AI has a long-term role to play.
For operators, they tipped the technology to transform many aspects of their businesses, as long as they are able to concurrently address concerns around privacy and security, ethical and social considerations.
And perhaps, the surest sign generative AI isn’t just another fad came at the recent 6G Symposium event I attended in London.
In discussing ongoing 6G development work, Vodafone Group’s senior R&D manager David Lister stated the sudden rise of generative AI systems made it harder to plan for the next-generation of mobile.
Lister noted it was difficult to assess the impact of platforms like ChatGPT on networks, both going forward and today, adding it is likely future work done on 6G use cases will include some related planning.
With 6G tipped to come to a phone near you not until at least 2030, it does indicate this technology is here to stay. And everyone, including operators, want a piece of it.
The editorial views expressed in this article are solely those of the author and will not necessarily reflect the views of the GSMA, its Members or Associate Members.Subscribe to our daily newsletter Back