> I don't think ChatGPT and other things are going to make that much of a difference since people have been plagiarizing and republishing crap content for ages.

I'm afraid you're being too optimistic here. True, the internet has been full of rehashed, google-bait "content", in fact there have been lamentations recently that search engines have became useless because of these (because most of the hits on the first (few) page(s) point to such less than useless "articles").

What ChatGPT et al. change is that they make creating such content almost effortlessly easy.

An illustrative recent anecdote from an adjacent field: Clarkesworld, a reputable science-fiction magazine, had to suspend evaluating submissions completely for a while, because they were overwhelmed by the sudden influx of bad AI submissions.

It's always been like this to some degree, and the internet is already full of crap content.

If you think it's bad for Perl and Python, you probably haven't worked with Arduinos much. You have to double- and tripple-check absolutely everything. Doesn't matter if it's beginner code, schematics or commercially produced circuit boards. Things like real time clock boards becoming a fire hazard to SD card modules monopolizing a multi-device SPI bus are too common. I won't even go into the mess that many I2C libraries internally initialize the bus, leading to problems when you use multiple of them in a single project...

> The trick is to not fight the spam.

I agree here, reacting to those blogs would be akin to feeding trolls. The added attention would help them.

> There's plenty of good stuff out there

yes!

> and it's easy to find if people try.

I'm not to sure about that, clever SEO is more and more dominating.

Just today I tried finding old threads of mine on Perlmonks via DuckDuckGo and got only hits on StackOverflow. (WTF?)

> However, the people most affected by bad sources

Mass matters and at some point the dams breaks , and not only out of laziness.

Of course I can tell people "don't trust the search engines", "ask me" or "read good books". But who will constantly spend 10min to 1h for the latter if the search engine is only seconds away?

And I agree with kikuchiyo here, that the economical aspect plays a role.

Producing nonsense also takes energy and time. But once it is cheap enough to use "AI" for that, we might end up in a sea of BS.

As an example:

Just look at how the quality ensuring mechanisms of "old media" were already bypassed by algorithms and trolls spamming "new media".

That was already possible with "old media" - like "letter to the editors", buying journalists or whole outlets - but is now far cheaper.