PieFedGoogle is experimentally replacing news headlines with AI clickbait nonsense
submitted by
Avid Amoeba34 Comments
Comments from other communities
Another reason to stay away from Google Discover.
RSS is the way to get news from online newspapers. It avoids the bias and nonsence injected by AI and algorithms when getting news via social media and big techs platforms.
P03 LockeDid you know that BG3 players exploit children? Are you aware that Qi2 slows older Pixels? If we wrote those misleading headlines, readers would rip us a new one
No, they wouldn’t. Those are the types of clickbait garbage headlines you use already.
coyotino [he/him]
luciole (he/him)And they’ll get AIs to scrape that shit too. And train on it so they can spew higher order nonsense. And so on.
Thanks for the archive link.
So what’s happening here is Google is feeding headlines into a model with the instructions to generate a title of exactly 4 words.
Every example is 4 words.
Why they think 4 words is enough to communicate meaningfully, I do not know.
The other thing is whether novel they’re shoving into their products for free is awful, hence the making things up and not knowing in the context of a video game exploit is not the same as the general use of the word.
I don’t think meaningful communication is a KPI they optinize for. More likely time spent in the Discover feed.
“Trump cry like baby”. Huh.
4 words fit better in the box or on the screen is what I assume.
Four words seems a fair metric for the shortest descriptive headline. I mean, can you beat “Foot Heads Arms Body”?
The only shorter ones are “man bites dog”, “Dewey defeats Truman”, or something as simple as “WAR” when everyone already knows the details and this is just the official announcement.
If hypothetically a false headline on a reputable site led to an incident involving injury or death, could Google be found liable in anyway?
No because on the google.com eula that you sign by having someone on your family ever Google something redeems them of any liability and gives them a right to sacrifice your first born to AI
EULAs are not legally enforceable anyways
They’re becoming closer and closer to it though. Scary court decisions are being made, it won’t be long before someone tests it as a legal argument
yet.
Liability waivers don’t apply outside the US.
Oh you didn’t see the clause that supercedes that? Silly consumer not reading everything.
/s
How are other countries enforcing that liability tho?
rarely
I doubt it.
Are you cooking something up?
They could hypothetically. Will they? Probably not.
Possibly. The BBC made a big row about Apple’s headlines: https://www.bbc.com/news/articles/cx2v778x85yo
didn’t this happen already? the thing is generating AI responses instead of showing me the results first and then I’m not clicking on it because I’m a person
it’s also de-listing a ton of websites and subpages of websites and continuing to scrape them with Gemini anyway
Apple had to turn it off for their sunmary mode after backlash, even though the option always had the “these summaries are generated by AI and can be inaccurate” warnings placed prominently.
Google doing this shit without warning or notice will get them in shit water. News portals and reporters are generally not too fond of their articles being completely misrepresented.
it’s not just a matter of misrepresentation. it’s directing traffic away from the websites which are creating the content, maybe depriving them of every means that they have of monetizing it
Well, the loss of traffic is a knock-on effect of the misrepresentation. So is the fact that every other portal will try to sling shit at the ones affected by it.
Clickbait and ai bullshit from Google feed is pretty much all I’ve ever seen from them in the past year.
Yeah my Ublock Origin block filter is nearly 2000 lines, almost all of it blocking shit like this. If it ever gets nuked on Firefox then I’m going full Kaczynski.
Great make every effort to suck the soul out of internet, of which there is already very little remaining. Then you can jack off to all the money you made while watching the internet that is %100 bots and racists
We’re just here on Lemmy
That part google is making will remain dead
How else are we gonna carry around a pack full of skeleton parts for our necrarmy
Wow yet another example of forcing AI into something that the AI just makes worse. Must be a day that ends in Y
It is Jeudi
I kind of like the idea of a system allowing me to automatically remove clickbait and sensationalism from headlines and replace it with a good summary. But I really hate how Google is pushing that without customization, without consent and in such a crappy state.
Same. I don’t mind it if it was done by a group of writers who can finetune it or people who just hate that shit. But but Google’s job is to get more clicks and get more eyeballs. So I don’t trust them to game their own systems.
i thought they were already doing that? idk i assume a lot of the news that gets read is AI generated. if you have a good prompter you can easy crank out a hundred thousand years worth of fake headlines.
They were generated by the news sites, not google itself