Inside a network of AI-generated newsletters targeting "small town America"

niemanlab.org

28 points·giuliomagnifico··4 comments

Pretty good idea! Especially now that SOTA tokens and email lists are so cheap. I don't understand why the tone of the article is so negative.

Well, this article is an example of journalism: somebody took an interest in some phenomenon and did some digging, made some phonecalls and compiled what they found into a narrative. This is outside the scope of LLM newsletters, and the article quotes people into why they find it distasteful (read: dishonesty)

Would it be that difficult for the LLM generated newsletter to disclose that it's a summary of information provided by humans?

The newsletters link to the sources. Clearly most people don’t want to read the whole article - especially if the linked source requires a subscription.

Decades ago there were people who didn’t really read the paper but just scanned the headlines - this is just a modern way of enabling that.

Still, it’s devious to monetize the existence of other content this way.

>Would it be that difficult for the LLM generated newsletter to disclose that it's a summary of information provided by humans?

A disclaimer like that would likely make many readers feel "less comfortable" I suppose!

When "AI-generated articles" are disclosed, it upsets people. For instance, see the experiment articles on TechCrunch, The Verge, or the similar websites (I don't remember exactly which one).

Reply on news.ycombinator.com