Yeah. Protectionist trade policies are going to be great for software engineers. What was a bad investment in automation software when it replaced a $1/hour job starts looking like a good investment when it's a $15/hour job it's automating away.
We've already seen this with jobs that are impossible to outsource, like retail. That self-checkout machine may have been expensive to develop, but it's a lot cheaper than employees. (The machines aren't perfect, but one employee can now check out 10 customers simultaneously.)
In some ways, nothing will change in America. The jobs were already gone, so nobody is going to lose their job to a robot. That said, it's time to start coming up with a real plan for unemployment. We aren't taxing automation. We aren't considering a world where labor demand is less than labor supply. We have no plan for the future, and the future is not going to be coal mining and making cheap flip flops for Wal-Mart. We are going to have to fund basic income, or some sort of welfare system without a social stigma. (All the foreign investment driving up housing prices isn't helping either.)
Of course, we aren't immune either. Most software isn't actually needed; as we improve off-the-shelf software, custom software will be in lower demand. (Ever pay anyone to write you a custom image manipulation program? Nope. Photoshop is fine. Really not sure why we aren't "done" with HR databases or shopping carts yet.) And, of course, machine learning will help fill in the one-off customizations that are sometimes necessary. And, maybe it's going to start creating art, music, literature, and whatever else we consider "uniquely human". Pretty sure I'll be dead by then, but someone should probably start planning. You thought "peak oil" was a problem... but we're already past "peek labor".
Which they won't be. The average American can't seem to understand that trade negotiations are just that, negotiations. Each side gives up something they would like to have for something they'd really like to have.
Additionally, protecting American labor isn't inherently good. Maybe these companies decide to bring back a few hundred $40k/yr manufacturing jobs, but offset the cost by outsourcing a few hundred $60k/yr office jobs.
We've already seen this with jobs that are impossible to outsource, like retail. That self-checkout machine may have been expensive to develop, but it's a lot cheaper than employees. (The machines aren't perfect, but one employee can now check out 10 customers simultaneously.)
In some ways, nothing will change in America. The jobs were already gone, so nobody is going to lose their job to a robot. That said, it's time to start coming up with a real plan for unemployment. We aren't taxing automation. We aren't considering a world where labor demand is less than labor supply. We have no plan for the future, and the future is not going to be coal mining and making cheap flip flops for Wal-Mart. We are going to have to fund basic income, or some sort of welfare system without a social stigma. (All the foreign investment driving up housing prices isn't helping either.)
Of course, we aren't immune either. Most software isn't actually needed; as we improve off-the-shelf software, custom software will be in lower demand. (Ever pay anyone to write you a custom image manipulation program? Nope. Photoshop is fine. Really not sure why we aren't "done" with HR databases or shopping carts yet.) And, of course, machine learning will help fill in the one-off customizations that are sometimes necessary. And, maybe it's going to start creating art, music, literature, and whatever else we consider "uniquely human". Pretty sure I'll be dead by then, but someone should probably start planning. You thought "peak oil" was a problem... but we're already past "peek labor".