You know how the machine learning systems are in the news (and Ted talks, tech blogs, etc.) lately over how they're becoming blackbox logic machines, creating feedback loops that amply things like racism on YouTube, for example. Well, what might the ML/AI systems be doing with our code repositories? Maybe not so much yet, I don't know. But let's imagine. Do you think it's probably less worrisome? At first I didn't see as much harm potential, there's not really racist code, terrorist code, or code that makes people violence prone (okay, not entirely true...), but if you imagine the possibility that someone might use code repositories to create applications that modify code, or is capable of making new programs, or just finding and squishing bugs in code algorithmically, well then you have a system that could arguably start to get a little out of control! What if in squashing code bugs it decides the most prevalent bugs are from code that takes user input (just one of potentially infinite examples). Remember though, it's a blackbox of sorts and this is just one of possibly millions of code patterns it's finding troublesome, and most importantly it's happening slowly (at first). Just like how these ML forces are changing Google and YouTube algorithms so slowly that many don't notice the changes; this would presumably be similar and so it may not be as obvious as one would think. So anyways, 'it' starts refactoring code that takes user input into something 'safer'. Great! But what does this mean? Not for this specific example really, but this concept of blackbox ML/AI solutions to problems we didn't realize we had, what does a future with this stuff look like (Matrix jokes aside)? Well, I could go on all day with imaginative ideas... But talking to myself isn't so productive, let's start a fun community discussion here! Join in if you find this topic as interesting as I do! :)

Note: if you decide to post something like "SNN have made this problem...", or other technical jargan please explain it as clearly as possible. As the great Richard Feynman once said, the best way to show you understand a thing is to be able to explain it clearly to others who don't understand it... Or something like that ;)