The iPhone ruined technology. 

There was an article in the paper the other day about Nissan Leaf owners who were angry because Nissan were discontinuing the app service that allowed them to connect with their cars and get updates on charging status amongst other things. In the grand scheme of things, this isn’t such a great loss. Most charging points will update you on charge complete time, and it’s also usually estimated in your car dashboard. At the end of the day, all these people can still drive their cars. This will not be the case for people with many other connected devices, whose providers have decided they should get a new whatever and so discontinue support for it, or even those whose providers go bankrupt. There are horror stories about people having systems fitted to restore their sight, only for the provider to go bust and leave them sightless with some dead technology implanted in their head. Most peoples’ experiences aren’t this dystopian, but the best outcome is your connected TV/food processor/oven just ends up as a very expensive paperweight. The fact is that some things in the world should not rely on a networked connection to a wider ecosystem, and yet many of them (including a worrying array of medical devices) now do. 

The Internet of Things (IoT) was a technological fad of mid 2010s, at the time it was touted as the future of technology, and companies rushed to connect everything to the internet. The result was a lot of internet capable washing machines whose processors were used for very little other than DDoS attacks. As this wave of ‘netification’ receded, the devices that actually benefitted from being connected to the internet (TVs, games consoles, music streaming devices) remained so and I’d like to say the ones that didn’t make any sense being connected to the internet went back to being standalone, but of course that’s not entirely the case. You can probably still buy a washing machine that is inextricably connected to the internet. I’m sure there are some people who will read this and say “I find being able to start my washing machine remotely very convenient when I forgot to start it or set the timer immediately after putting the washing in it,” but my understanding of the genuine utility of that is pretty well defined in the statement itself. 

Many of these things persist because they’re there now and as long as no one actively complains about them (regardless of whether they use them) what’s the harm in leaving them? If this is a form of evolution, it is a particularly crappy one: survival of the not actively annoying. Unfortunately it is a pattern that has repeated itself again and again with each wave of technological ‘progress’. Blockchain was going to change everything – contracts were going to be immutable, art was going to be non fungible, currency was going to be international – but in the end it turned out to just be an easy way to take money off marks and bribe a president. Then of course the metaverse was going to replace most of society and work. Wild predictions were made by the sorts of people who make predictions about tech and finance, seemingly based entirely on how much money Mark Zuckerberg was willing to spend on its development. As a measure of how far that broke containment from the tech/finance world, when I told a friend the other day that Meta was shutting down Horizon Worlds, its main metaverse space, his response was “what’s the metaverse?” The thing is, there is much that thrives in the metaverse, but it’s things like Roblox or Minecraft rather than some sort of grand network of virtual spaces, or whatever it was that prompted Meta to invest a total of $82 billion in a project that appears all but over with the end of Horizon Worlds. 

We should take some time to think about that number, because at slightly more than the GDP of Venezuela it is a hell of a lot of money for no actual return. However, it’s not a lot of money to buy the future. I don’t mean a stake in the future, I mean all of it. In the same way that he’s always wanted to own all of social media but never quite managed it, Mark Zuckerberg wanted to own the whole of the future, or at least all of the infrastructure on which the future will run. His bet was that this was the metaverse. This is the dream of all the tech leaders, why do you think Elon Musk branded X ‘The Everything App’? He, at least, believes it literally can be. This is also the logic behind generative AI. Companies like OpenAI and Anthropic believe that if they can replace enough of the functions that other software and platforms do, their ‘Al’ will become the medium through which we interact with the world. Obviously that would require some pretty dramatic shifts in the capabilities of LLMs (by which I mean them basically working in a completely different way) to actually work, but don’t let that stop you investing another few billion dollars in an LLM major based on what computers can do in Star Trek. 

The point is that none of these companies is actually trying to innovate. They are not trying to build useful technology, they are trying to build indispensable technology, whether it is useful or not is immaterial. Why though? Why with all the hundreds of billions of dollars they have available don’t they want to do better? Greed possibly – if you create a monopoly then you get all the income – but I think that’s only part of it. There is also a lack of imagination, but it’s not a lack of imagination about what exciting new technological paradigms they could be inventing*, it’s the lack of imagination about how a technology company could be successful without being a monopoly. It’s not their fault, they had bad examples to follow, or at least they had misleading ones. The internet changed everything about how we communicate, the mp3 changed everything about how we consume music and the mobile phone changed everything about how we behave in public, but only the iPhone managed to take all these things and lock them up in a closed, monetise-able ecosystem that people actually wanted to be in. It also spawned a large secondary industry of app development as companies saw the potential of a new window on the world (or a novel version of the old one). From then on everyone was on the hunt for the next technology-as-ecosystem, either because they wanted to own it or build their business on top of it. In this way we moved from a world where people identify problems and then build technology to solve them to a world where people identify technology and then look for problems to solve with it, almost regardless of whether it’s up to the task. A lot of money and effort is poured into making sure everyone thinks that the latest thing discovered (because these things are rarely completely new technology) is the new thing. A specific kind of new thing: not the thing that will be the new internet or the new World Wide Web (although it is commercially sensible to claim inheritance from these things too), but the thing that will be the new iPhone.  If the claim is convincing enough, everyone will try to find a use for it and whatever this new tech is gets stuffed into your fridge, your TV and your attempts to engage with your fellow humans. Whether it remains in those things depends on a) whether it is perceived not to lose money** and b) whether it can avoid annoying enough users that it becomes a PR liability.  Both of these things are really about the potential stock price but I’m not going to get into that here. 

As the tide of hype recedes, leaving behind a flotsam of new and rehashed functionality, no one takes the time to evaluate what the real benefit of that new technology was. Unnecessary functionality is baked into technology, increasing tech debt whilst adding very little real consumer advantage. If these services can be used for charging higher fees they will be retained and maintained regardless of their actual ROI (or real world use). Equally if they have no ongoing revenue potential, they can be switched off at a moment’s notice leaving customers who used those services high and dry (à la Nissan Leaf drivers). No one considers the effort and money expended on trying to make the last fad fit in their product and whether it was worth it, usually because they’ve already moved on to the next fad. The costs are ‘development costs’, investments in innovation that have value as signifiers of a company’s growth potential, basically marketing spend. 

For those of us who work in the (frankly) less sexy parts of the technology sector, an alternative approach is necessary. Client requirements or clear potential ROI are prerequisite for any development, so it’s baffling for us to understand how so much fluff seems to get cludged into other technology. At least until you consider the other thing Apple introduced to the world of technology shortly before the iPhone: the idea that technology could be cool. The recent announcement that Tim Cook would be stepping down as CEO of Apple, to be replaced by another identikit tall white man in a black T-shirt reminded me of this other driver of tech adoption. Apple had always relied on a strong brand identity, presenting themselves as a culture more than a company. Of course before the age of the iPod, that culture was a counterculture. Mac users were the underdogs, the people who used Macs as a protest against the bland corporate orthodoxy of IBM, Microsoft, et al (and, as those users would argue, because they were better). In the early 00s Apple leaned into the countercultural cache that their somewhat fanatical customers gave them. Technology as lifestyle gradually became almost standard, expanding the fanaticism of the core Apple customers into a kind of weird identify politics of technology. You were an Apple person or an Android person, often citing fractional privacy advantages or novel functional elements (usually quickly adopted universally) as justifications for your meaningless decision to identify with one techno tribe or another. In retrospect it’s quite embarrassing really. Obsessing about minuscule (and largely inconsequential) changes in technology had traditionally been the preserve of nerds. Strictly speaking, it still is, but there are just a lot more nerds these days. Being, at least superficially, interested in technology became normal and to a certain extent even cool. 

Such a situation lends credibility to those who claim greater knowledge of the ‘secrets’ of technological progress and gives birth to a whole industry of techno palm-readers. Not just the consultants paid handsomely to foist new technology on organisations and then move on before its efficacy can be fully assessed, but increasingly a whole gang of cranks and snake oil salespeople with only a basic grasp of the fundamentals of the technology but an uncanny ability to generate FOMO. This simply accelerates the take up of a new technological fad because it’s ‘cool’ rather than because it has a generalisable utility. Generative AI has taken all this to new heights of proto-religious fervour, not least because much of the current mania is based on wild assertions that machines are about to become conscious, based on no actual evidence at all. The difference between this fad and all others is anthropomorphism. Because the technology can plausibly replicate our language, we are more keen to ascribe intention, intelligence and ability to it. We appear happier to believe it is capable of things regardless of the lack of evidence of its capability. Generative AI is possibly the first technology to become its own snake oil salesperson. Unfortunately this means the number of ill advised integrations of generative AI is exponentially greater than for any previous technological fad. On top of that, so many of these integrations are of the Nissan Leaf app variety, in that they rely on the supply of a third party service that has absolutely no guarantee of persisting. I realise that if one LLM service goes down, it could be replaced by another, but there’s no guarantee of exact equivalence. That is just another way in which this technology strives to be more iPhone: once a service is gone, it’s gone. We should all bear that in mind as we rush toward the next fad. 

*although possibly also that.

**in the era of the meme stock, whether a company actually makes money or not has become almost immaterial.