Telecoms Glom Onto ‘AI’ Hype Cycle In A Fruitless Bid To Make 5G Seem Interesting

We’ve noted for several years how the “race to 5G” was largely just hype by telecoms and hardware vendors eager to sell more gear and justify high U.S. mobile data prices. While 5G does provide faster, more resilient, and lower latency networks, it’s more of an evolution than a revolution.

But that’s not what telecom giants like Verizon, T-Mobile, and AT&T promised. Both routinely promised that 5G would change the way we live and work, usher forth the smart cities of tomorrow, and even revolutionize the way we treat cancer. None of those things wound up being true (I enjoyed talking to one medical professional who basically laughed in my face about the cancer claim).

When 5G did arrive, it didn’t even live up to its basic promise, really. U.S. implementations were decidedly slower, spottier, and more expensive than many overseas networks, thanks to the usual industry consolidation and U.S. regulatory fecklessness. The end result: wireless carriers associated a promising but not world-changing technological improvement with hype and bluster in the mind of consumers.

With the ink barely dry on the disappointment, telecom providers are now trying to suggest that “AI” (read: language learning models and machine learning) is just the ticket to “rescue” 5G from irrelevance in dramatic fashion. Verizon, for example, has struck a new partnership with NVIDIA it claims will supercharge excitement over 5G and stalled telecom edge computing efforts all at once:

“Our ongoing investment in our network infrastructure means we’re uniquely positioned to deliver these powerful AI services at scale, driving the digital transformation and fueling the future growth of businesses worldwide.”

Telecom analysts have noted that deploying some additional AI compute resources in the radio network (AI-RAN) might bring about some efficiency improvements, but just like 5G itself it’s more iterative than transformative. As always, innovation-stifled telecoms want to be seen as innovative, key players in the AI and edge computing markets, but they’re usually not, notes Dean Bubley:

“Telcos have demonstrated only a minimal role in edge computing services, either as localised low-latency cloud computing suppliers, or even in terms of just offering colocation space in exchanges, or mobile towers / aggregation sites.”

AT&T’s tried similar things, like this 2023 announcement of a partnership with NVDIA the companies promised would “supercharge operations,” “enhance experiences for both our employees and customers,” and “build, customize and deploy interactive avatars that see, perceive, intelligently converse and provide recommendations to enhance the customer service experience.”

In the same announcement, AT&T pat itself on the back for the company’s climate change and energy efficiency initiatives, ignoring (or trying to pre-empt criticism of) the massive power costs of AI.

The press coverage of these announcements always winds up rather bubbly and unskepctical. AT&T, it should be noted, continues to have some of the worst customer service ratings of any company or industry in America, which is no small feat when you consider that health insurance, medical care, and airlines exist. U.S. 5G is still among the slowest in all developed nations.

Telecoms operate in a market that doesn’t get much hype or press attention in the “Big Tech,” crypto, and AI era. In part because network management isn’t all that sexy or hugely profitable. But also because they largely operate in minimally competitive fields rife with regulatory capture where they’re not really incentivized to truly innovate.

So to drive some market and press interest they’ll desperately try to offer “me too” -esque services and shallowly jump on board of hype trains to latch on to some of the money in other fields they’re envious of (it’s really how the net neutrality fight started after AT&T declared it would double dip on Google way back in 2002).

South Korean telecom giant SK Telecom, for example, was all about the Metaverse when it was the hyped new hot thing. Now it’s pivoting seamlessly pivoted to trying to pretend it’s a cutting edge AI company. Once the AI hype dies down some they’re inevitably glom on to some other technology they’ll pretend they’re at the cutting edge of. It’s just how this pattern goes.

Which wouldn’t be quite so bad for telecoms here in the States if their core competencies weren’t so shaky, with U.S. broadband and wireless still some of the spottiest, slowest and expensive in the developed world, with shaky customer service to match. You hear endless chatter about technological innovations in telecom, but the actual consumer experience always winds up a yard short.