I'm not necessarily talking about us becoming machines/cyborgs ourselves, I can't even imagine that happening (but not doubting it either). That's a little too magical thinking for me right now.
What I do believe: more so that machines will be making a lot of decisions with much more accuracy than we can when given the same data a human is.
8 Ways Machine Learning Is Improving Companies’ Work Processes
Because decisions are based on data and experience at the end of the day. A machine can store data and observe the conclusions, then store that data and make comparisons. It's not new, but it's getting a lot more automated. Even the human elements/interactions are all data at the end of the day.
There's a human element, there always is, but there aren't many who call shots anyways. If a shot caller sees it fit that employing a machine gets better results and is cheaper, they'll cut people off. Machines learning capabilities have taken of very recently.
The problem will be when many people have been cut off. Then that's where things like UBI come into play or face total rebellion.
Also there doesn't need to be a big distribution of hardware, programmers tend to plug into machine learning API's. Like Google's new TensorFlow CPU doesn't have plans to distribute, but you can make calls to their API to use it for your own or company's personal projects.
One of my recent projects I was going to see if I could grab data from an NBA API and use one of Google machine learning api's with it.
The Great Strengths and Important Limitations Of Google's Machine Learning Chip
Plus with open source development having such a recent impact where everyone can be a part of the development opens the doors for a level of group thinking that was never possible and efficient as before.