I bought an iPhone 15 the other day to replace my five-year-old iPhone 11. The phone is powered by the new A17 Pro chip and has a terabyte of data storage and accordingly was eye-wateringly expensive. I had, of course, finely honed rationales for splashing out on such a scale. I’ve always had a policy of writing only about kit that I buy with my own money (no freebies from tech companies), for example. The fancy A17 processor is needed to run the new “AI” stuff that Apple is promising to launch soon; the phone has a significantly better camera than my old handset had – which matters (to me) because my Substack blog goes out three times a week and I provide a new photograph for each edition; and, finally, a friend whose ancient iPhone is on its last legs might appreciate an iPhone 11 in good nick.
But these are rationalisations rather than solid justifications. The truth is that my old iPhone was fine for the job. Sure, it would need a new battery in time, but apart from that it had years more life in it. And if you take a cold, detached look at the evolution of the iPhone product line, what you see from the 2010 iPhone 4 onwards is really just a sequence of steady incremental improvements. What was so special about that model? Mostly this: it had a front-facing camera, which opened up the world of selfies, video chat, social media and all the other accoutrements of our networked world. But from then on, it was just incremental changes and price rises all the way.
And this is true not just for iPhones but for smartphones, generally. Samsung, Huawei, Google and other manufacturers have been on the same path. The arrival of the smartphone signalled by the launch of the first iPhone in 2007 represented a sharp discontinuity in the evolution of mobile phone technology. (If in doubt just ask Nokia or BlackBerry.) There then followed a massive surge for a decade or so, until the technology (and the market) matured and incremental changes became the rule.
Mathematicians have a name for this process. They call it a Sigmoid function and draw it as an S-shaped curve. When you apply it to consumer electronic devices, the curve looks like an “S” that’s been flattened a bit. Progress is slow at the bottom; then it takes a sharp upward turn, before eventually flattening out at the top. And smartphones are now on that part of the curve.
If we look back at the history of the tech industry over the last five decades or so, we can discern a pattern. First there’s a technical breakthrough: the silicon chip; the internet; the web; the mobile phone; cloud computing; the smartphone. Each breakthrough is followed by a period of frantic development (often accompanied by investment bubbles), which propels the technology up the middle bit of the “S”; and then eventually things calm down as markets become saturated and radical improvements in the tech become more and more difficult to achieve.
You can perhaps see where this is heading: to so-called “AI”. It’s already had its initial breakthroughs: first, the arrival of “big data” produced by the web, social media and surveillance capitalism; then the rediscovery of powerful algorithms (neural networks), followed by the invention of the “transformer” deep-learning architecture in 2017; and then the development of large language models (LLMs) and other forms of generative AI of which ChatGPT was the poster-child.
We’ve now had a period of frenzied development and insane amounts of corporate investment (with no clear idea of what the returns on that investment will be), which has propelled the technology up the central part of the sigmoid curve. So interesting questions now arise: how far up the sigmoid curve has the industry climbed thus far? And when will it reach the plateau, where smartphone technology currently reclines?
In recent weeks we’ve begun to see signs that that moment may be approaching. The technology is being commoditised. The AI companies have started to release smaller and (allegedly) cheaper LLMs. They won’t admit this, of course, but that may have something to do with the way the energy costs of the technology are ballooning. The industry’s irrational boosterism cuts little ice with economists. And while millions of people have tried ChatGPT and its peers, most of them haven’t displayed enduring interest. Virtually every large company on the planet has had an AI “pilot” project or two, but few of them seem to have made it into actual deployment. So could it be that this sensation du jour is about to get boring? A bit like the latest shiny smartphone, in fact.
What I’ve been reading
Zero sum games
A transcript of a remarkable talk by Maciej Cegłowski, one of digital technology’s sharpest observers, on the moral economy of tech.
In the frame
Vivian Maier: reclusive nanny, great street photographer, subject of a lovely essay by Ellen Wexler in Smithsonian magazine.
Baby bomb
Ed West’s sobering review of Paul Morland’s book on the world’s coming demographic crisis.
