And 5-10 years ago self-driving trucks were going to put truckers out of a job in 5-10 years. Turns out the world is complex.

I think it's much easier to replace digital workers because

1. Significantly more people are tackling digital problems (eg. you don't have many solo entrepreneurs working on self-driving car projects because of the high capital requirements, regulations, and it being fairly niche)

2. Replacing digital work is safer because unlike driving it generally doesn't involve life or death.

How about this. If I’m wrong I owe you a Coke.

Governments are not capable of reacting to this within five years. It’s barely an election cycle. Predicting that all countries will have introduced basic income within five years is rather bold.

Anyway, there’ll be more programming work (because it’ll be cheaper) before there’s less programming work (because it’ll mostly get automated away). It might even pay better.

I think UBI will be the norm by 2030. Though the way I see it playing out is that initially most countries will just opt for stronger safety nets with means-testing. The point at which workers' livelihoods are clearly being automated in mass is where it will start to become politically feasible to implement the UBI. I think UBI is too radical for the baby boomers, but more palatable for the youth who've been massively screwed over their whole lives and are more aware of how rigged the system is.

Mediocre Programmers Will Be Mostly Obsolete in 5 Years

Human programmers will be considered mediocre compared to machines in 5-10 years, just like chess and Go.

Ironically, if true, the net effect of displacing programmers will be lower demand for AI because because of the resulting economic collapse.

I'm still waiting for the "all money will become useless in 5 years due to Bitcoin"!