Mobile World Congress Curtain-raiser: Feb 2024

 

Article by The Minister of Truth

The Mobile World Congress in Barcelona is promising to return with an even bigger bang at the end of February as it gets back up to speed following the fallow Covid years.

 

And this year, it promises to be bursting with innovations, many of which will have an AI dimension. Because when it comes to communications, there are few areas that are not impacted by AI these days, from network infrastructure and management, to the content that streams over those networks, from the data centres that process this burgeoning mountain of data, to the latest smart devices promising to put the power of superintelligent computing in our pockets.

 

Along with the usual suspects from the telecoms industry, it’s notable that one of the keynote speakers is Demis Hassabis, CEO and Co-founder of Google Deepmind – a figure as ubiquitous as his AI software these days – who will undoubtedly be plugging Gemini, Google’s largest and smartest AI model to date.

 

So buckle up and prepare for a thrilling ride along the AI highway.

 

You might have thought full 5G, or 5G Stand Alone (5G SA) – which doesn’t rely on the older, less efficient 4G network and so can offer much higher speeds and capacity – has barely left the traps. And you’d be right. Industry observers have been surprised at just how slowly network operators have been launching their 5G SA networks around the world.

 

But they’re already excitedly talking about 6G. I kid you not. Some experts at MWC24 will be speculating that AI won’t be able to reach its full potential without 6G (sixth-generation wireless), which promises even higher frequencies, even lower latency (delay), and even higher capacity than 5G SA.

 

These superpowers will enable 6G networks to serve many more people with much faster throughputs of data. And thanks to the higher frequencies they’re likely to use, 6G networks, with the help of machine learning and AI, could even be used as sensing tools. Interference in the patterns of radio signals could reveal how many people are in a building at any one time, for example, enabling traffic monitoring and improving energy management systems. But that’s future-gazing – we’re not likely to see this until 2030.

 

So why is this relevant to you and me?

 

We’ve talked quite a bit already about Generative AI (GenAI) on this site, and its ability to generate convincing images, video, audio and text synthesised from all the existing content the internet has to offer. It promises to usher in an era of hyper-personalisation, involving chatbots that really get to know you and marketing material that is much more targeted and relevant, for example.

 

This means more data. Lots more data. No really, absolutely massive amounts of extra data. Why? Well, as we know, image files are always much larger than text files (roughly five to ten times bigger, depending on the level of compression, pixel ratios and so on). Video files tend to be even bigger.

 

Now imagine what they call volumetric, immersive content – 3D scenarios that you can actually walk around and underneath wearing specialist wireless headsets like Apple’s new Vision Pro. These files are likely to be humungous in size and will need superfast, AI-optimised, networks to handle them. Which brings us back to 5G SA and 6G.

 

Before, content creators would call on image libraries - or commission their own photos and videos - to populate their websites and marketing emails. Now they can generate images and videos themselves very easily and quickly using GenAI. And the unshakeable golden rule of innovation is that if you make something easier to do, more people will do it.

 

So this democratisation of content creation is going to lead to the aforementioned exponential growth in digital data, data that has to reside somewhere on racks of servers powered by electricity. And that comes at a cost.

 

Data centres worldwide already devour as much energy as the state of California, and some US states are requesting that the planned phase-out of coal and gas plants is delayed to accommodate massive new data centres being built by the major cloud computing providers, such as Microsoft.

 

But the truth is we don’t really know what impact all this extra data and computing power is going to have because we don’t have reliable figures for data centre energy usage. Optimists say that energy efficiency improvements and continued growth in renewable energy supply will mitigate against this surge in demand; pessimists worry that the AI-fuelled data surge will accelerate climate change Armageddon.

 

At least the European Union is trying to regularise reporting. Its new Energy Efficiency Directive requires all but the smallest data centres in the bloc to report on their energy use and sustainability. Which has to be a good thing.

 

The challenges presented by more computing power, more data, more energy lead to a logical question. What if we didn’t have to rely so much on cloud computing and giant energy-hungry data centres to enjoy the full benefits of AI? What if we could bring things closer to home – to the ‘edge’ as the techno-nerds call it.

 

This brings us to the next exciting development that will undoubtedly receive lots of attention at MWC24 – localisation. This just means building AI into your device’s chipset, so you can access it faster, making more use of your smartphone or laptop’s prodigious, and often underappreciated, computing power. This is a hot topic in the tech world right now and all the chip and device manufacturers are working on releasing the next-gen fleet of gadgets that will bring GenAI right where you need it, without having to worry about flaky network connections. Apple argues that this will have privacy benefits, too.

 

This localisation trend will go hand-in-hand with the specialisation of Large Language Models like ChatGPT, so that they can be applied to specific data sets that companies can trust and control, and tailored to company needs.

 

Ian Fogg, tech analyst at research company CCS insight, sums it all up like this:

 

“AI has been around for some years, used for specific cases. But now it’s being applied in many more areas - network, devices, software - such that every tool we use now has the potential to become much, much better because of AI.

 

“And AI has the potential to make networks greener, and the world a more efficient place.”

 

 Mobile World Congress 2024 looks set to be a fascinating event.

 
Guest User