As a programmer and computer scientist my relationship with AI is...well, complicated. On the one hand AI is poised to turn this particular human profession into a mere hobby, something for those of us predisposed with the inclination. On the other, there is no denying the fascination with the possibilities, the realization of something that for my whole intellectual life had been a science-fictive dream.
But, if i’m honest, it’s more than that.
For the last many years i had become ever more disenchanted with the actual business of programming, often spending my workday looking for the correct, if arbitrary, incantation that would unlock some predefined functionality in yet another framework or reiterating some variation of a solution to a (usually trivial but necessary) problem i’d solved many times before.
Once i started playing with LLMs, the currently most advanced form of AI we have, i realized that all of that, all those frustrations of essentially mindless repetition and yak-shaving, the boilerplate for the panoply of opinionated languages and their host of pre-written functionality, all that was no longer my responsibility. And good riddance.
So, if you are somehow irreconcilably opposed to the concept that AI can or should produce useful software, even if guided by a knowing and experienced mind (for now), then you will not be curious about the projects below. Be warned, though, for the future is set and there is no way of knowing, henceforth, the provenance of any artefact in existence.