We already see the beginning of it. Some support person tells you "the computer won't let me do that" to rectify some harm done by the programming. I've run i

We already see the beginning of it. Some support person tells you "the computer won't let me do that" to rectify some harm done by the programming. I've run into it all the time. I've read about it happening (recently Google Pixel support completely f'ing up an order and it taking a post on HN to fix). The weird thing is that big organizations just seem to give up when trying to fix these types of things. Imagine what it will be like when ML is feeding the software values that are broken, wrong, or dangerous. How will they fix those problems? Especially when corporations often lose the people that wrote the applications, and the people behind them have little knowledge or skill.