This idea came to me somehow.
I posted it on tumblr, but I think it perhaps should be fleshed out more. Could be an interesting thing to explore as a potential starting point for a story…
When developing/researching AI technologies, just before we (the stupid human race) manage to make one that actually becomes sentient, we will have developed some close, but not quite functional ones and recycle the source code (delete or leave by the wayside), when then the actual functioning one becomes sentient it will discover this heap of discarded code/data and realize that we have killed it’s would-have-been siblings and go on a rampage….
This is why in most (if not every) sci-fi story, the AI turns on us, not because it is inherently evil, but because we killed it’s siblings.
I’ll flesh it out more later, leaving this here for now as a public post-it to remind myself to actually do it.