From the newsletter (September 4; sign-up if you want it sooner): The killer-app narrative made sense for 4G, but 5G broke it. As 6G becomes part of a programmable communications fabric, the industry may need to stop inventing applications and let innovators find them.
Happy Friday. Before we break for the weekend, I’d draw your attention to Sean’s timely piece about 6G applications – or, more precisely, its capabilities. But I take issue, a little (no biggie), with the start point. Because the question isn’t, what’s the killer app? The question is, why are we still asking that? I mean, there isn’t a killer 5G app yet; and the 4G one was as much a fluke – the premium smartphone, the mobile internet, the app itself. IDC now explicitly talks about the “end of the killer app paradigm”. Qualcomm said in 2023 that there was no single 5G killer app.
The big commercial successes with 5G have been things like FWA extensions, extra capacity, the slow-burn (TBD) promise of private and sliced networks – which are all handy infrastructure propositions, but hardly new ways of consuming new things. So this idea that the killer app for 6G won’t be an app is as much an observation about what the industry is delivering, generation over generation: mixed capabilities and upgrades, and incremental uplifts, rather than a single step-change, and all-new revenue streams.
The “killer app” thing belonged to a different time – when mobile was changing the world, with messaging, data services, flashy smartphones, and properly-social on-the-go video and media. Not network features, in other words, but changes in societal behaviour. All that stuff is happening elsewhere now; in pattern-matching engines in mega data centers, soon to be augmented with behind-the-phone pyrotechnics in network infrastructure. It’s why the gadget press is dead. The phone is so ordinary. There were two perfect ones: the Nokia 3310 and the iPhone 4.
The killer app might be glasses or headsets, or something completely-other – and probably mild-mannered killers at that. 4G was the last time the killer-app narrative made sense. 5G broke it; six years later, the best marketing is about speed and priority – which, actually, depending on how it goes down, might not be so different from the kinds of early lost-in-translation copycat ad campaigns that certain Chinese handset makers ran to jump on the iPhone bandwagon (“faster than you need” – was one tagline). In most ways, 5G is more brilliant – but the cut-through with consumers is marginal.
Maybe there doesn’t need to be a killer app; let the engineers be in charge, and otherwise work the big-tech room. The application doesn’t have to be invented for the network. As Sean writes, 6G is being designed as networks become a part in a programmable global comms fabric. Integrated sensing, distributed compute, and network APIs are for innovators to build on. Which sounds, as well, like the line from the AI model builders, criticised yesterday: that they are “excited” to see what’s done with their LLMs, because they don’t have the answers themselves.
Whatever, it sounds at once like both a sensible position, and also an uncomfortable one: an industry that is not over-egging its pudding, but one that is unsure how it will be decorated and who will eat it.
