• Get involved.
    We want your input!
    Apply for Membership and join the conversations about everything related to broadcasting.

    After we receive your registration, a moderator will review it. After your registration is approved, you will be permitted to post.
    If you use a disposable or false email address, your registration will be rejected.

    After your membership is approved, please take a minute to tell us a little bit about yourself.
    https://www.radiodiscussions.com/forums/introduce-yourself.1088/

    Thanks in advance and have fun!
    RadioDiscussions Administrators

False and Apparently "AI" Derived News Story about KNX's FM changing Format in LA.

Status
Not open for further replies.
You can post boogieman alerts all day. The fact remains that much of this has little to do with AI that's all up in the news today. News Bots have been around for years. Much of the traditional press is just jumping on the AI bandwagon because they don't understand what it is, other than they've heard it makes crap up. Over the years have there been bot-generated stories? Sure, that's what news bots were invented for. Scrape other news sites and social media, then compile it into a story that comes off as salacious. Are news agencies interested in AI because it could save them money over employing reporters to surf the Internet looking for news? Sure, but not if the AI gives them a bad reputation. This assumption that AI is being suddenly released to the new-consuming public is, creating its own narratives. Which, is what news sites want actually.
 
AI is simply not a credible source of news or information -- they simply scrape together information from multiple sources to put together something that sounds like it might be plausible, but there doesn't seem to be any real fact checking as part of the AI algorithms. Out of curiosity a couple months back, I decided to ask ChatGPT for a history of the "TM Stereo Rock" automated radio format that was big in the seventies and early eighties. The result that I got back showed not even a minimal attempt at accuracy.

There may be some things that AI is well-suited for, but providing fact-checked information of any sort is not currently one of those things.
 
michael hagerty said:
No. There's a maximum number of hours they can broadcast separately and still qualify for single-line reporting in Nielsen. Probably originally intended to cover situations like sports coverage on one station, but not another and things like that, but Audacity uses it to make a few bucks on the AM while keeping the news rolling on the FM.
Click to expand...
Yes, this is a loophole in the Nielsen requirement for "almost 100% simulcast" which was implemented to cover stations with play-by-play sports contracts that did not allow one of the simulcast partners to be used for the broadcast.

A typical case, at the time, was an FM that simulcast with a big AM. The FM covered the metro, but the AM covered a larger area, overlapping other sports team affiliates. So the station could not simulcast on AM. The Nielsen rule was made to allow for the total run time of "a couple of games".

KNX used that exception to keep the paid junk on weekend mornings while still being listed as a full simulcast. The rule was not intended to cover paid programs on one signal, but the situation does comply with the rule.
 
Kelly A said:
You can post boogieman alerts all day. The fact remains that much of this has little to do with AI that's all up in the news today. News Bots have been around for years. Much of the traditional press is just jumping on the AI bandwagon because they don't understand what it is, other than they've heard it makes crap up.
Click to expand...
I don't think that Granny's misunderstanding of AI has spread to people in "the press" who at least know that it means some kind of "computerized search and consolidation of common threads that appear to be widespread enough to be actual facts or ideas". They may not understand how few of these are reviewed by sentient beings that are not plugged into an AC outlet.
Kelly A said:
Over the years have there been bot-generated stories? Sure, that's what news bots were invented for. Scrape other news sites and social media, then compile it into a story that comes off as salacious.
Click to expand...
However, AI only works when there is a mass of information out there so that common sets of words and expressions and phrases are found, matched and assembled. The more junk online, the more chance of finding an inaccurate non-fact and assuming it is actually correct and real.
Kelly A said:
Are news agencies interested in AI because it could save them money over employing reporters to surf the Internet looking for news? Sure, but not if the AI gives them a bad reputation.
Click to expand...
But with news sources having trouble being profitable, and without deep research and verification at traditional press outlets, the chances of "fake news" being widely circulated increase. All four of the largest US traditional news sources, the NYT, WaPo, LA Times and the WSJ, have all cut back the newsroom recently. The only daily paper in Santa Barbara dropped dead last week. And so on. Less reputable, responsible news media members, more operations that are struggling to make ends meet.
Kelly A said:
This assumption that AI is being suddenly released to the new-consuming public is, creating its own narratives. Which, is what news sites want actually.
Click to expand...
But it is actually true.

I started my own website 20 years ago due to the wealth of misinformation on social media sites and places like Wikipedia and find that the misinformation about everything related to radio and its related fields in technology and music seems to expand exponentially, somewhat like Moore's Law in its growth.
 
CTListener said:
I'd say something about typical German efficiency here, but it would probably be misinterpreted as either insulting to Germans or sympathetic to the Nazis. Not good. I might just write a catchy country song in protest, "Try That in a Beer Hall."
Click to expand...
My German story, from some years back:

Walking along the riverside in Dusseldorf, we came upon an open air bar. Decided to have a beer, which seemed to be right for the environment. Ordered two. No, only beer mit schnaps. OK, charge me for the schnaps, but just give me the beer. No, you have to take the schnaps as we don't sell beer alone. Whatever. Drank the beer, enjoyed the flavor of the moment and learned a bit about Germans.

Left the schnaps to our neighbors at the bar.
 
DavidEduardo said:
I don't think that Granny's misunderstanding of AI has spread to people in "the press" who at least know that it means some kind of "computerized search and consolidation of common threads that appear to be widespread enough to be actual facts or ideas". They may not understand how few of these are reviewed by sentient beings that are not plugged into an AC outlet.
Click to expand...
I sort of agree, but Granny is just seeing that AI is taking over-something via the news. That combined with her memories of watching the Terminator movies, and she's seeing AI everywhere.
DavidEduardo said:
However, AI only works when there is a mass of information out there so that common sets of words and expressions and phrases are found, matched and assembled. The more junk online, the more chance of finding an inaccurate non-fact and assuming it is actually correct and real.y
Click to expand...
It depends. Some Large Language models like ChatGPT/Microsoft Sidney-Bing/and Google Bard, all do the same thing; they search portions of the public Internet looking for something to respond with that is accurate to what it's able to find, and if not, make something up that would satisfy what it thinks the questioner would consider the response reasonable. In other words, the model is designed to please.
DavidEduardo said:
But with news sources having trouble being profitable, and without deep research and verification at traditional press outlets, the chances of "fake news" being widely circulated increase. All four of the largest US traditional news sources, the NYT, WaPo, LA Times and the WSJ, have all cut back the newsroom recently.
Click to expand...
They have indeed, but just like radio, I don't see a day where legitimate newpaper publishers would allow scraping of the Internet by AI, News Bots or whatever, just to save money. A few instances of publishing false information that damages credibility would be much more expensive than a couple fact checkers/editors. The same way radio isn't going to just stick some budget AI algorithm hooked to a voice synthesizer to just rattle off crap it finds on line.
DavidEduardo said:
The only daily paper in Santa Barbara dropped dead last week. And so on. Less reputable, responsible news media members, more operations that are struggling to make ends meet.
Click to expand...
Santa Barbara is a pretty high-end community with younger families already online. I can see where similar communities in that socioeconomic environment would be in the same boat.
DavidEduardo said:
I started my own website 20 years ago due to the wealth of misinformation on social media sites and places like Wikipedia and find that the misinformation about everything related to radio and its related fields in technology and music seems to expand exponentially, somewhat like Moore's Law in its growth.
Click to expand...
Sure, but you've had to fine-tune that information over time as it becomes available in order to get the historical record correct. Most modern news consumers don't spend the time to check, or double check what they consume on line. Especially if the story aligns with their personal narrative, or crosses the 'sexy' bar.
 
Kelly A said:
I sort of agree, but Granny is just seeing that AI is taking over-something via the news. That combined with her memories of watching the Terminator movies, and she's seeing AI everywhere.
Click to expand...
No, Granny and her younger-but-still-Luddite friends just have a narrower perspective and see fewer shades of grey. But she sees "red" and "blue" way too clearly.
Kelly A said:
It depends. Some Large Language models like ChatGPT/Microsoft Sidney-Bing/and Google Bard, all do the same thing; they search portions of the public Internet looking for something to respond with that is accurate to what it's able to find, and if not, make something up that would satisfy what it thinks the questioner would consider the response reasonable. In other words, the model is designed to please.
Click to expand...
It also takes what appear to be common online statements to be true if they predominate. Obviously, the algorithm is more complex, or we'd see more "Trump won" articles...
Kelly A said:
They have indeed, but just like radio, I don't see a day where legitimate newpaper publishers would allow scraping of the Internet by AI, News Bots or whatever, just to save money.
Click to expand...
In many small markets, it is happening now. Some newspapers are using local TV station news coverage to do breaking news articles, and reducing the investigative reporting to nothing.
Kelly A said:
A few instances of publishing false information that damages credibility would be much more expensive than a couple fact checkers/editors. The same way radio isn't going to just stick some budget AI algorithm hooked to a voice synthesizer to just rattle off crap it finds on line.
Click to expand...
Yet...
Kelly A said:
Santa Barbara is a pretty high-end community with younger families already online. I can see where similar communities in that socioeconomic environment would be in the same boat.
Click to expand...
Santa Barbara, if you do not consider the university population as "permanent" is almost as old as the Palm Springs MSA. It should have lots of traditional newspaper users and lots of people interested in local news.

(Part of the issue there was bad management by an "iconic" rich woman who owned the paper but ran out of the money she was willing to spend. We can't depend on philanthropy to provide news coverage)
Kelly A said:
Sure, but you've had to fine-tune that information over time as it becomes available in order to get the historical record correct.
Click to expand...
No, I don't "fine tune" anything. I just present the printed documents in books, magazines, journals and other formats that were considered reputable at the time. No interpretation at all.
Kelly A said:
Most modern news consumers don't spend the time to check, or double check what they consume on line. Especially if the story aligns with their personal narrative, or crosses the 'sexy' bar.
Click to expand...
Absolutely. The problem is that they don't, in their vast majority, understand how much wrong information is readily accessible on the web. For a while, I kinda' tracked radio related Wikipedia articles to see what percentage had at least one major inaccuracy. Those ranged from "big picture" subjects like Major Armstrong and Marconi to individual radio stations or on-air personalities. After finding the "gotcha" to be about 80%, I quit that and scanned some more documents.
 
DavidEduardo said:
No, Granny and her younger-but-still-Luddite friends just have a narrower perspective and see fewer shades of grey. But she sees "red" and "blue" way too clearly.
Click to expand...
And we see that here on this very site. Many things appear as binary: You have good 6+ ratings, so a station is considered the winner. Anything less is a loser. AI is the cause for 'fake news' online. Even though false stores have existed online since the public Internet started. They don't remember AOL Newsgroups, or Compuserve discussion boards.
DavidEduardo said:
It also takes what appear to be common online statements to be true if they predominate. Obviously, the algorithm is more complex, or we'd see more "Trump won" articles...
Click to expand...
That depends. Some of the new models are considered 'Constitutional' AI. That means there are constitutional constraints baked-in to avoid. Taking a political side is a common guardrail.
DavidEduardo said:
In many small markets, it is happening now. Some newspapers are using local TV station news coverage to do breaking news articles, and reducing the investigative reporting to nothing.
Click to expand...
Many news radio stations do that too. At least that sort of reporting is better than scraping the public Internet.
DavidEduardo said:
(Part of the issue there was bad management by an "iconic" rich woman who owned the paper but ran out of the money she was willing to spend. We can't depend on philanthropy to provide news coverage)
Click to expand...
Just like owning a local radio station, owning a local news periodical is a thankless role. You zig, and certain members of the public complain about some form of bias one way. You zag, and it happens from another group. Try and straddle the middle, and you're considered boring. It's akin to owning a sports team. Everyone thinks the team is their personal sports team, and you're messing it up.
DavidEduardo said:
No, I don't "fine tune" anything. I just present the printed documents in books, magazines, journals and other formats that were considered reputable at the time. No interpretation at all.
Click to expand...
So you're saying you rely completely on Wikipedia info, and don't seek out information by scanning information which satisfies that you're publishing correct info? And if so, that's not considered fine-tuning?
DavidEduardo said:
Absolutely. The problem is that they don't, in their vast majority, understand how much wrong information is readily accessible on the web. For a while, I kinda' tracked radio related Wikipedia articles to see what percentage had at least one major inaccuracy. Those ranged from "big picture" subjects like Major Armstrong and Marconi to individual radio stations or on-air personalities. After finding the "gotcha" to be about 80%, I quit that and scanned some more documents.
Click to expand...
And again, once you determined the discrepancies, you sought out corrections or clarifications? Not considered fine-tuning?
 
Kelly A said:
You zig, and certain members of the public complain about some form of bias one way. You zag, and it happens from another group.
Click to expand...
Solved by buying a pack of ZigZag and sitting back to fully chill.
 
Status
Not open for further replies.


Back
Top Bottom