The hype/scare re GPT-2/3 (etc.) is not for their poetry output, but rather for its potential for mass propaganda, telemarketing and so on. We can already get humans to do this stuff, all GPT could give is scale (that's no small deal).
However, if the output needs to be curated and edited by humans, the scale and automation is gone - we just get a different manual process, with a modest improvement to speed at cost of some decline in quality, and that's not very impactful.
The truly scary part is SEO where GPT-3 could ruin search engines overnight.
Google at this point favours long form content for many search intents. Being able to generate thousands of these pages in one-click is a real problem. Not just because of popular topics e.g. "covid-19 symptoms" but more so for the long tail e.g. "should I drink coffee to cure covid-19".
Quite a lot of SEO already uses simple word generation techniques. It isn't clear GPT-3 is an improvement there - human text recognition might not be whatever Google does.
It may be that Google's algorithms don't care at all how human-like the text is, or that their own recognition algorithm/NN (whatever they use) isn't fooled. Even if it is affected, Google has the money and corpus to build its own competing NN to recognize GPT-3 text.
While I have no doubts that they could build NN capable of recognizing GPT-3 text I believe that this would still pose a problem given the amount of content to be analyzed at the scale that Google deals with
I'm sure Google out of all entities could handle scale.
That said, there might be a different threat to Google. GPT-3 seems really useful as a search engine of sorts (with the first answer implementing the 'I'm Feeling Lucky' button). Tune it for a query syntax, and for getting the 'top X' results somehow, then we just need the web corpus and a basic filter over the results. We could have a very interesting Google competitor.
However, if the output needs to be curated and edited by humans, the scale and automation is gone - we just get a different manual process, with a modest improvement to speed at cost of some decline in quality, and that's not very impactful.