Yes, it is perfectly ‘valid’ to be concerned. But I would ask you to think about this:
Yes, there are a lot of online articles — some with some rather scary-looking statistics — that some people consider very convincing. But the vast majority of the people who read these articles are not economists who are trained in statistical analysis techniques. Making an article that can ‘seem reasonable’ (or even ‘convincing’) to a layperson is a relatively easy task. But someone trained in Economics and Econometrics might look at the same article and not be convinced by it at all — because they are trained to be able to detect dodgy economics and misinterpreted statistics.
When economists study these things seriously, they obviously don’t just look at a few scary-looking graphs. They collect vast quantities of data and subject that data to some pretty rigorous computer-based regression analysis. If they could put together a paper that would convince their fellow economists that an automation crisis was on its way, I expect we’d hear about it. Those economists might even be on their way to a Nobel Prize.
Now, we can base our judgements on the impressions of laypeople scouring the internet for articles and statistics that sound impressive to them. But I suggest that listening to the conclusions of skilled economists might be a much better option.
And if people don’t want to trust the economists, there’s always the option of studying Economics themselves, starting with a nice and simple Introduction to Economics book.
Thanks for adding your thoughts, by the way! :)