The LA Times has finally activated its long-awaited, AI-powered "bias meter," announced last December by its owner, Dr. Patrick Soon-Shiong. It appears at the bottom of all opinion pieces and looks like this:
Color me a little unimpressed. Practically everything was labeled center left, and that's all it does. Meh.
Just below is a "Perspectives" dropdown, which provides a dull summary of the piece, along with an even duller set of bullet points offering alternative viewpoints. I don't really see the point of that. I'd prefer a fact check. That might get the fur flying.
All this stuff is generated directly by AI. No human staff sees it before it goes up.
Next version: hotdog, or not hotdog?
Seems we might not want AI doing a "fact check" until the hallucination problem (a.k.a. "alternative facts") is solved.
facts of any kind are not allowed in magaworld
We live in an age where any idea that can’t be characterized as “center left” is more than likely batshit crazy.
I was just thinking the same thing. At this point, even the first Trump administration feels center left.
+2
LA Times seems to be giving up on print. It's day-after-Oscars coverage every year has been a mainstay for as long as I remember. Front-page stories, a separate section, and more. Lots for readers all over town on the biggest news day of the year for the local biz.
Today, not a single mention of the Oscars in print except for a QR code link to the online edition of the paper.
Except for Michael Hiltzik, not sure why to subscribe anymore.
I stopped. I subscribe to the SF Chronicle for California news. The Mercury-News has a lot of bay area news, but the Chronicle covers more of the state, I think.
And where's their Overton window?
More interesting would be an originality or novelty meter...the only way AI would be fairly accurate with the bias meter is by tracking speech consistent with known platitudes ("in the box"), and then that begs the question how much original and/or novel thinking is going on at all.
The truth has a well-know center left bias.
Rational and reasonable: tl:dr "center left"
An AI bias meter seems totally pointless. Anyone who pays attention should be able to place an article's contents on the political spectrum. If you are so clueless that a bias meter provides useful information, the knowledge that a bias exists is hardly sufficient for making an informed opinion - it cannot help you determine fact OR personal values, both of which are central to deciding your own position on an issue.
If AI could provide immediate citations and context to all the arguments presented in a piece, that might be useful. A clueless person reading about Zelenskyy and Trump's meeting could benefit from information about Ukraine's political history from the October Revolution to now, Putin's history of meddling and military action, Trump's impeachment, the people who currently influence Trump, etc. But doing that would likely be seen as far too biased for the LA Times, and it is unclear that AI is up to the task or that the clueless would ever read it.
Agree and well said. I read WaPo (at least for now) and they've started to summarize comments via AI. It's almost insulting the reader's intelligence. I suspect there's an element of "here's what you should think" for people who will be inclined to read the summary and skip the rest of the article or comments.
This, as well as those suggested responses to previous comments, seem geared to reducing actual representation, not just eliminating simple up-votes & down-votes.
And their online "ask" option is annoying; i prefer to simply see the options - print, opinions, comics, etc. & select what I want. People whine about the "nanny state," but feeling as though the Bezos pick to run his paper wants to direct readers, even to asking for letters to the editor on specific topics, as if subscribers to The Post need to be encouraged or directed, is unnecessary. And no, I am not inclined to use the AI option that they suggest I need. I'll wait for a while to see what's worthwhile.
Why the AI bias meter? Oh, that one's easy: Soon-Shiong wants eat as much of the LA Times former reputation as he can before it becomes yet another right wing horse head crawling with eels. Or maybe lampreys.
what a worthless waste of time and stupid use of AI. the news organization shouldn't be labeling articles based on an arbitrary viewpoint score, what they should be concerned about and what readers should want is accuracy. what is even the point of this?
Right now it's just featured with opinion pieces, but I'm sure that's just a beta test. Soon-Shiong wants to eventually apply it to the whole paper in a quixotic attempt to appease MAGA. Which of course it won't because you can't appease Nazis. You'll notice the Epoch Times isn't reciprocating with its own AI meters. For some reason.
It's too bad -- for a while the LA Times was doing yeoman's work covering California and the nation *without* being in a defensive crouch politically and columnists like Hiltzik and Lopez are national treasures. Guess those days are over.
Center Left. In America...
[Ray Liotta laughing gif]
I took a look at the inspection panel and I don't see any code there. Was that on purpose?
LAT celebrates false equivalence with a dumb meter.
I'm with kahner: what any and every news organization should be striving to do, and should tout to their readers, is the accuracy of their reporting. How they put the "stories" into context. Analysis that goes beyond the momentary fads and "memes" of the moment. Getting the whole "story", not just the click grabber aspect. Oh, and NOT having a pre-determined editorially-set "narrative" that everything had to be fitted into (and parts that didn't, left out.)
I remember, several years back, that NPR announced a policy like this, and to their credit, it lasted a couple of weeks.
I think that word does not mean what he thinks it means