Automation is a given, more-and-more things are going to be automated. People jobs are going to be the last to be automated too, perhaps these jobs never will

Automation is a given, more-and-more things are going to be automated. People jobs are going to be the last to be automated too, perhaps these jobs never will be automated, although a lot of low-level traditionally people-facing jobs will, e.g. shop assistants, it's already started. But notwithstanding the above, there are logical holes in the rest of the article. I won't even mention that slight on high-tech work by equating it to "plugging away at a spreadsheet". The biggest flaw was the assumption that the automation to make the aforementioned spreadsheet obsolete would be done without those people. In reality, any person highly-paid to "plug away" at a spreadsheet would be paid due to their domain/industry knowledge and skills rather than purely Excel skills and would still be employed operating a new system, or even directly involved in building and maintaining such systems. This pattern is repeated throughout the article. Technology will obsolete low-tech grunt work first, mainly because contrary to the central premise of that article, technology is a people business. It's not a people business in the same was sales or marketing is, it's not about who's got the biggest set of contacts[0], but you'll never build a successful product without understanding the world it fits in to; building good technology requires a good deal of listening and empathy, both toward the customer, but also to the system as a whole and your fellow engineers. In the short-to-medium term, any acceleration in automation will also increase demand for engineers in that field. And any such systems will only be as good as those who made them, even knowledge-workers not in the automation business will still be required to operate/design/improve such systems. Maybe in smaller numbers... What happens after that, no one knows. The effects on the economy would be difficult to predict. And one day in ten, fifty, a hundred, several thousand years from now, fully autonomous AI will be here... and all bets are off. Long-term predictions would be foolish. [0] - except for the sale and marketing of a technology business, of course, it's not a binary situation.