[kodah]

I don't really think the world needs AI right now. One can argue that the AI is making an innocent mistake and that calling an AI or ML (or it's improper training, however that works) "racist" is overblown rhetoric as people are here, but I think all of that aschews the actual issue. The problem is that AI and ML are primarily used for decision making, like in recommendation engines. These little gadgets that provide recommendations may be fairly low-stakes, but are theoretically proof-of-concepts for future applications like policing, fighting terrorism, or human trafficking. If you get it wrong there, the consequences are devastating. If people don't raise the flag about how wildly wrong the AI is now, then there will inevitably be a false confidence to use it for the aforementioned applications (and there are plenty of examples of how this has already happened).