Google AI Overview made up an elaborate story about me

>The 6pt text at the bottom clearly says:

I did inspect element and it's actually 12px (or 9pt). For context the rest of the text (non-header) is 18px. That seems fine to me? It's small to be unobtrusive, but not exactly invisible either.

You are right. It is okay to do whatever you want, as long as there is a sign stating it might happen.

Especially in an area you own, like your own website or property.

Want to dump toxic waste in your backyard? Just put up a sign so your neighbors know, then if they stick around it's on them, really, no right to complain.

Want to brake-check the person behind you on the highway? Bumper sticker that says "this vehicle may stop unexpectedly". Wow, just like that you're legally off the hook!

Want to hack someone's computer and steal all their files? Just put a disclaimer on the bottom of your website letting them know that by visiting the site they've given you permission to do so.

>You are right. It is okay to do whatever you want, as long as there is a sign stating it might happen.

Stop strawmanning. Just because I support google AI answers with a disclaimer, doesn't mean I think a disclaimer is a carte blanche to do literally anything.

These are great life hacks! Thanks for sharing

Reading your comments in context of the thread you're on, you think the disclaimer is sufficient to do things up to and including falsely claiming public figures have opposite views to their true ones on the Israel-Gaza conflict.

Considering the extent to which people have very strong opinions about "their" side in the conflict, to the point of committing violent acts especially when feeling betrayed, I don't think spreading this particular piece of disinformation is any less potentially dangerous than the things I listed.

Where do you draw the line then? I doubt the AI is assessing the risk of 'what happens if I fuck this up', so perhaps the feature should be removed?

So why GP reasoning doesn't apply to Google AI snippets and you consider it a straw-man? Classic search results are clearly not Google's, they just match (or not) with your search query, then you go there and read them (and trust them or not depending on your own criteria or absence of). But a text, generated by Google, put as the first paragraph of text under your search, answering in plain English to a specific question you just asked, what should a disclaimer like that supposed to be? A "read it but discard it because it could be factually wrong"? Why are they showing it topmost?

I do understand it is a complicated matter, but looks like Google just want to be there, no matter what, in the GenAI race. How much will it take for those snippets to be sponsored content? They are marketing them as the first thing a Google user should read.

>Classic search results are clearly not Google's, they just match (or not) with your search query, then you go there and read them (and trust them or not depending on your own criteria or absence of).

What you said might be true in the early days of google, but google clearly doesn't do exact word matches anymore. There's quite a lot of fuzzy matches going on, which means there's arguably some editorializing going on. This might be relevant if someone was searching for "john smith rapist" and got back results for him sexually harassing someone. It might even be phrased in such a way that makes it sound like he was a rapist, eg. "florida man accused of sexually...". Moreover even before AI results, I've seen enough people say "google says..." in reference to search results that it's questionable to claim that people think non-AI search results aren't by google.

Fun fact along this line of reasoning: all those dump trucks with the "not responsible for broken windshields" stickers? Yes, yes they are responsible.

You can't just put up a sticker premeditating your property damage and then it'd a-okay.

No, the sticker is there to deter YOU from suing in small claims court. Because you think you can't. But you can! And their insurance can cover it!

Since you are clearly an AI enjoyer I asked my local LLM to summarize your feelings for me. It said:

> As evidenced by the quote "I think a disclaimer is a carte blanche to do literally anything", the hackernews user <gruez> is clearly of the opinion that it is indeed ok to do whatever you want, as long is there is a sign stating it might happen.

* This text was summarized by the SpaceNugget LLM and may contain errors, and thusly no one can ever be held accountable for any mistakes herein.