I'm less optimistic on that front. I think there is a lot of potential for positive effects coming from ML, but I'm not so sure that the potential negative eff

I'm less optimistic on that front. I think there is a lot of potential for positive effects coming from ML, but I'm not so sure that the potential negative effects are less likely and/or lesser in magnitude than the positives. I think I would be more optimistic if I hadn't witnessed the internet go from a bastion of hope for mankind to a smoldering heap of attention robbing waste. The two are similar in that it isn't the technology, it is how we utilize it that determines whether the outcome is good. One particularly concerning aspect of ML is the lack of understanding on how the model reaches the 'correct' answer. This aspect of the technology encourages people to 'let go' and rely on the answer without understanding anything about the context. The trouble is that we embed context while training the models, so the context is there, we just can't see it once it is trained-in. It will be an interesting decade, for sure.