There's something profoundly disrespectful about feeding snippets produced by human authors en masse into a machine learning algorithm to create a predictive text generator to be applied as labor replacement, regardless of accuracy or appropriateness to the context or use case. There's something profoundly disrespectful about forcibly scraping the human corpora without asking for permission first, or considering compensation in anyway, or just setting out that "yeah, we're doing this, and damn the consequences". There's something profoundly disrespectful about then taking this device running on investor money chasing the promise of a value that just can't conceivably be delivered, and employing it to do all the unethical things you could imagine just because you can engineer the "No" out of the model.

Of course the level of moral disengagement that's rife around here generally precludes any recognition of such things, as that might be construed as "harshing the vibe on the innovation, man". So I guess everyone had better just settle in and get used to some serious profound disrespect until this type of thing hits enough of a critical mass that a societal ethics framework begins to manifest around it and not be ignored.