As 5G gets rolled out across the globe, consumers’ demands are also set to climb to a level not seen in past generations. They are already doing so: according to an analysis by Opensignal in 2020, 5G users spent more than double the mobile data compared to 4G users.

That means up to 38.1 Gb per month in South Korea, a nation where consumers have become early heavy users of 5G.

While the demand for larger bandwidths leads to the need for more robust technologies able to manage the amount of data flowing, experts argue it also allows for these same technologies to evolve. That is the case with artificial intelligence and edge computing.

“When I see 5G and edge computing I’m trying to see what the opportunity is,” said Michael Liou, VP Strategy & Growth at Chooch AI, in an interview with 6GWorld. According to him, some technologies already function in an environment lacking even 3G, like detecting fires on-premise and sending the data back via radio. But when 5G appears, it takes the game to a different level.

“If we had 5G [at such places], we’d have much more streaming, we could send much more data. We could actually send things without as much of a delay, and the richer the information the better someone can react,” the specialist explained.

According to him, 5G makes virtualization more seamless and more frictionless as well. That is when “lines start blurring between what is virtualised and what is actually on-premise.”

“There’s a lot of situations where we need edge computing because of the remote connectivity issue or the immediacy or the critical nature of what we’re trying to do. And 5G helps bridge that gap.”

The “gap” Liou is talking about also refers to enabling other technologies. He pointed out that the workload and deployment of AI, for example, depend upon latency, reliability, and scalability issues.

It works like a chain, in which the next step relies on the previous one to happen. With bandwidth, reliability, and technologies like AI and edge in place, telecoms and users can start exploring the boundaries of the 5G ecosystem, coming up with new solutions for recurring problems that demand heavy data flow.

“The nice thing about cloud is that you have almost infinite scale. I can put as many computer-vision models on as many GPS as you’d like on your private or public cloud. And then [we start asking] how much data are you sending? What’s the transfer cost? This is where 5G comes into play. If it’s relatively cheap and you can have the workload on the cloud and a very robust 5G connection, then that just might be able to solve most of your issues.”

AI and 5G Applied to Real Life

What will our everyday lives look like when we have all the connectivity up and running to the max?

Maybe you and I won’t notice a visible change – but it will be there, taking place in both small and big moments of our day.

“You can install a home security camera [at your front door] for $200. But regardless of what happens on your front door, [the notification you receive] usually just says ’movement at your front door.’ It doesn’t give you any information. It could be a false positive because maybe the tree cast a shadow, or a car moved in the background,” Liou explained.

“What happens if the information is ‘A package has been delivered’, ‘UPS has delivered a package’, ‘Amazon has delivered a package’, or ‘Three of your family members showed up?’ How much more valuable would that be?”

Liou pointed out that while the current technology already allows companies to make advances in this specific area, many are working on the various privacy issues brought by solutions like AI and facial recognition.

Protecting Data Is Crucial

If telecom providers are still figuring out the real power of 5G and scratching the surface of its capabilities when coupled with AI and edge, they are also concerned about protecting the insane amount of data generated. The entire ecosystem is, actually, because the consequences can be disastrous.

In November 2021, for example, British facial recognition company Clearview AI was fined £17 million ($22.6 million) for collecting billions of photos from sites like LinkedIn, Facebook, and Instagram to build up its database.

Enterprises are now figuring out ways to make sure personal data remains invisible. The solution Chooch AI found was something similar to tokenisation.

“If we were to take a picture of you and create a model, it’d be done in less than a second. Your face is very unique, and it would generate a unique mathematical value that would analyse 512 points on your face,” Liou explained. “We don’t need to know who you are. It’s just a unique number.”

That single number could be passed along to another company, which could then create a solution for its portfolio based on the person’s behaviour without ever compromising their privacy.

While it is not quite clear yet what 6G will look like in ten years – when it is projected to at least be standardised – there are lessons we can already learn from the privacy challenges with 5G.

“It allows the further deployment of AI and computer vision. Telecoms get to partner with us and with other infrastructure providers. We can update our AI on their boxes to provide better service via telecoms to end users. It’s pretty cool.”