People these days are discussing small language models that can be run locally on personal computers, which might be the future. There is a caveat to that, and there are lessons for us from the software development era of Microsoft and Google.

Software runs in a partially observable and changing environment. We discovered that operating systems are a necessary infrastructure to decouple resource allocation from the application itself. That means for application developers, there is a platform police to run past. Users’ affinity to spend time updating software is never as much as developers want. Building desktop software means fewer and fewer deployments. Leading to slower bug discovery, security, and performance issues, slower iterations from experience. Backward compatibility becomes paramount and hard to achieve in runtime that developers can not see. All those mean higher development costs.

The application and OS need an intermediate platform. A good and stable web browser on a common standard solved platform integrations, security, and packaging. The application developers do not have to worry about the OS version compatibility. Process-level optimization and memory was completely abstracted out for the first time, and events became first class objects without really taking any additional complexity. Even today, it’s significantly harder to develop for macOS compared to other apps on the Google Chrome browser.

Decentralized access matters more than decentralized execution. There is a great operational advantage in simply running a web server and connecting to a remote interface. It’s harder to run workloads and computations frequently, going beyond the personal computer’s capabilities. Sometimes, reliable technology requires centralization, such as in instant messaging. Oddly, because of the policies needed to support security in decentralized email required for it’s partial centralization. It is harder to ask people to pay for the software before significantly proving its worth. That led to a pricing model where cheaper services were increasingly available on the internet. This is one of those times, historically, when access matters more than privacy.

People always want a better toy. We now have more compute on watches than my personal computer from high school, and the computer on the Apollo mission. I have to make a philosophical leap to make the argument. People will never run out of work. The tree of imaginative and physical abstractions of nature always grows. There’s an unbelievable compositional ability among the objects in nature and the abstract, on which the market trades forever. Unless some shit happens. All this means, as long as consumers upgrade hardware slower than companies can pool significantly larger compute and offer a better toy, building locally is not an obvious choice.