13. The iPhone Moment: From Incubation to Growth
The future is not to be forecast, but created.1
It can takes years, decades even, for new computing platforms to go from prototype to mainstream adoption. This is true for hardware-based computers like PCs, mobile phones, and VR headsets, and it’s also true for software-based virtual computers like blockchains and AI systems. After years of false starts, someone releases a breakthrough product that kicks off a period of exponential growth.
The PC industry followed this pattern. The Altair was one of the first PCs2 when it launched in 1974, but the launch of the IBM PC3 in 1981 kicked off the industry’s growth phase. Even then, enthusiasts mostly used PCs to make games and hack around. Incumbent computer companies dismissed PCs as overpriced toys because they didn’t solve problems for their existing customers who wanted higher-end machines. But then PC developers built applications like word processors and spreadsheets,4 and the market exploded.
The internet developed this way too. The incubation phase took place5 in the 1980s and early 1990s, when it was mostly a text-based tool used by academia and government. Then, with the release of the Mosaic web browser in 19936 and the wave of commercialization that followed, the growth phase kicked off and has continued ever since.
AI has had the longest gestation period of any computing movement to date. The researchers Warren McCulloch and Walter Pitts7 conceived of neural networks, the core models that underpin modern AI, in a 1943 paper. Seven years later, Alan Turing wrote his famous paper outlining what people now call the Turing test,8 the idea that truly intelligent AI could answer questions in a way that is indistinguishable from a human. After many so-called summer and winter cycles, when funding sources came and went, AI now appears to be going mainstream, eighty years from its inception. Advances in graphics processing units,9 or GPUs, the special computer chips that underpin the technology, are a major reason why. GPU performance has been improving along an exponential curve, enabling neural networks to scale to trillions of parameters, the key driver of the intelligence of AI systems.
When I was an entrepreneur and starting out as a part-time investor, around the time the iPhone debuted in 2007, everyone was talking about mobile computing. My friends and I were starting to explore potential mobile applications and everyone wanted to know what the “killer apps” might be. Recent history provided a clue. Some apps that were already popular on PCs would likely translate to mobile, it was safe to assume. Shopping and social networking would no doubt continue to be popular adaptations. These mobile apps would be the skeuomorphic uses, taking existing activities and making them better.
Another clue came from mobile’s novel capabilities. Killer apps seemed likely to take advantage of these unique traits. The iPhone had many things that PCs didn’t. The device was always with you. It had GPS sensors and a built-in camera. These features enabled native uses, brand-new things you couldn’t do before.
In retrospect, the biggest hits closely followed this pattern. Breakout apps exploited the unique capabilities of mobile phones while also reimagining popular activities. Instagram and TikTok were social networks that relied on the camera. Uber and DoorDash were on-demand delivery services that relied on GPS. WhatsApp and Snapchat were messaging apps that relied on always being with you.
In 2007, the big question for mobile was, what kinds of mobile apps would matter? Today the big question for blockchains is, what kinds of blockchain networks will matter? Blockchain infrastructure only recently matured enough to support internet-scale applications. The industry is likely now nearing the end of its incubation phase and entering its growth phase. It is a good time to be asking what a killer blockchain network might look like.
Some blockchain networks will be skeuomorphic, doing what could have been done before, but better. Social networks are an obvious choice. They’re where people spend the most time, they influence billions of users’ ideas and behaviors, and they’re the primary economic engine for creators. Blockchains can create social networks that eliminate the high take rates and capricious rules that characterize today’s corporate networks.
Another important skeuomorphic category will likely be financial networks. Sending money should be as easy as sending text messages. Improving how payments work is mostly a collective action problem, something blockchains are well suited to solve. Blockchain-based payment systems could lower fees, reduce friction, and unlock new categories of applications.
There will also be important blockchain networks that are native, doing what couldn’t be done before. I expect many of these will involve media and creative activities. Other native applications will intersect with emerging areas like AI and virtual worlds, as I discuss in the sections ahead.
Inevitably, there will be categories of applications not covered here that end up being important. Entrepreneurs and developers who build the future are always going to outsmart armchair predictions. Nevertheless, I will try to make some informed guesses about what popular blockchain networks we might see in the read-write-own era. The list is non-exhaustive. I hope it will get you thinking.