I can see one avenue in which GPT-3 might be quite disruptive, and that is in reinventing our software development processes. For example, if it can be purposed into something that reliably converts source code between programming languages, then there is no longer any moat in library code and bindings beyond the prompt development; PL development will accordingly accelerate. And the same goes for tasks like developing tests, static checks, optimizations, user interfaces, import/export and so forth.

In this scenario, software itself commoditizes to a greater degree. Perhaps not to the point where the AI is a Star Trek computer, but transitionally towards that. And that means that software businesses increasingly become commission shops that pump out AI-built programs on demand, while software services built on platform lock-in get threatened by cheap data export and format conversion tools.