CNET blacklisted by Wikipedia

[BachChain]

Summary

 

CNET, once considered one of the major technology-focused news outlets, has been heavily criticized for falling editorial standards since it's acquisition by Red Venture in 2020. After years of dealing with AI-generated articles and other "advertiser-driven" decisions, Wikipedia's editors have had enough and are removing CNET from their list of reliable publishers and are categorizing any article published since November 2022 as unreliable. Some editors are going further and pushing to have content from any outlet owned by Red Venture automatically marked as unreliable

https://en.wikipedia.org/wiki/Wikipedia:Reliable_sources/Perennial_sources

 

Quotes

Quote

"Let's take a step back and consider what we've witnessed here," a Wikipedia editor who goes by the name "bloodofox" chimed in. "CNET generated a bunch of content with AI, listed some of it as written by people (!), claimed it was all edited and vetted by people, and then, after getting caught, issued some 'corrections' followed by attacks on the journalists that reported on it," they added, alluding to the time that CNET's then-Editor-in-Chief Connie Guglielmo — who now serves as Red Ventures' "Senior Vice President of AI Edit Strategy" — disparagingly referred to journalists who covered CNET's AI debacle as "some writers... I won't call them reporters."

Quote

"It's high time for it. Enough is enough," a familiar voice, bloodofox, concurred, adding that "if it's owned by Red Ventures, we need to go ahead and identify it as a hard [Reliable Sources] fail."

My thoughts

Enshittification claims another one. Journalism is the absolute last place that AI should be deployed to.

 

Sources

https://futurism.com/wikipedia-cnet-unreliable-ai

[PowerPCFan]

1 hour ago, BachChain said:

Summary

 

CNET, once considered one of the major technology-focused news outlets, has been heavily criticized for falling editorial standards since it's acquisition by Red Venture in 2020. After years of dealing with AI-generated articles and other "advertiser-driven" decisions, Wikipedia's editors have had enough and are removing CNET from their list of reliable publishers and are categorizing any article published since November 2022 as unreliable. Some editors are going further and pushing to have content from any outlet owned by Red Venture automatically marked as unreliable

https://en.wikipedia.org/wiki/Wikipedia:Reliable_sources/Perennial_sources

 

Quotes

My thoughts

Enshittification claims another one. Journalism is the absolute last place that AI should be deployed to.

 

Sources

https://futurism.com/wikipedia-cnet-unreliable-ai

AI is helpful, but at least:

1. Put a warning at the top of the article "This is AI-generated"

2. Have HUMAN editors fact-check articles

3. Just admit that you're using AI to write articles, don't pretend you're not using it when you are. (This problem is fixed with Step 1.)

 

It's still experimental - no AI is perfect. For example, the Google Gemini issues lately - or OpenAI Sora - all AI makes mistakes and every chatbot, language model, and generative AI is still experimental. Last week, I was testing AI and [I forget if it was ChatGPT or Google Gemini] told me that an i7-2600 from 2011 is better than an i3-13100f. So I'm fine with people using AI, as long as a human corrects its mistakes. 

 

And I'm OK with affiliate links, as long as they are like this:

 

Link to example product (affiliate link): example.com/affiliate-link-123

[LAwLz]

48 minutes ago, PowerPCFan said:

It's still experimental - no AI is perfect.

I think it's also worth mentioning that humans can be really shitty journalists too. 

Some websites not being recommended to use a sources on Wikipedia is not a new thing. It was a thing long before AI. The problem here is in my opinion not that AI is being used. It's that quantity is being prioritized over quality, and CNET not bothering to make sure the content they publish is of high standards. If your articles contain a lot of factual errors then you are not a reliable source, regardless of who or what wrote the article. If the AI didn't make mistakes then it would be reliable. A human making lots of factual errors would also be an unreliable source. 

 

 

Speaking of reliable sources, futurism.com is a pretty sketchy site that is highly opinionated and has a strong anti-AI agenda. In fact, if you go to the talk page for this change you can see several of the Wikipedia staff saying that futurism is a bad source. 

I think it's kind of funny that they are celebrating getting CNET removed from the list of reliable sources when they themselves aren't seen as reliable. 

[AlwaysFSX]

3 hours ago, BachChain said:

has been heavily criticized for falling editorial standards since it's acquisition by Red Venture in 2020

Now hang on a minute. I remember them making plenty of trashy articles in the early 2010s even.

Now people are taking notice because they're having AI generated articles?

[Kisai]

10 hours ago, AlwaysFSX said:

Now hang on a minute. I remember them making plenty of trashy articles in the early 2010s even.

Now people are taking notice because they're having AI generated articles?

If you think C|Net is bad, you should see the absolute garbage that's been produced ever since... oh the adpocalypse on youtube.

 

All these garbage articles, with ads on them, drive garbage traffic and rapidly devalue the advertising as real advertisers no longer want to be associated with garbage.

 

And if you think that doesn't happen, look at nearly any UGC site (including sites like slashdot and fark) where the value of the ads is so poor that the site has become 50% ads. Now think about these crappy AI generated ad farms getting their articles in front of UGC news aggregation like slashdot or fark. Now suddenly a site that was still functional becomes so utterly full of garbage that the site itself looks like AI generated garbage.

 

That is rapidly the direction things are going if some kind of regulation on what qualifies as "news" or AI labeling doesn't become a defacto requirement.

 

But AI is ultimately going to kill ad-supported services pretty quickly at the rate it's going, Why would you ever pay for programatic ad placement when that just means it will end up on some crappy AI generated article. So large advertisers would go back to manual ad buys, and if your site isn't as big as something like eBay, then your site dies.

 

[FlyingPotato_is_taken]

I have more and more issues with AI generated websites.

 

Does somebody know if there is a ublock list for AI content frams? I am sick of looking something up just to notice after a few seconds that it's AI gibberish. 

[jagdtigger]

14 hours ago, LAwLz said:

I think it's also worth mentioning that humans can be really shitty journalists too. 

Sure they can, but they can be good as well. AI in contrast is just plain bad all the time......

[denco101]

13 minutes ago, jagdtigger said:

Sure they can, but they can be good as well. AI in contrast is just plain bad all the time......

Plus, if one looks at the article in the OP, this wasn't the only issue that they had with CNET either.

[Forbidden Wafer]

Wikipedia should blacklist themselves then...

[starsmine]

2 hours ago, Forbidden Wafer said:

Wikipedia should blacklist themselves then...

Ill bite.
Why? 
Do they not strictly follow editorial standards and regularly correct themselves?

They should not be used as a source because of circular sourcing, not because wiki is wrong

[Sauron]

16 hours ago, LAwLz said:

If your articles contain a lot of factual errors then you are not a reliable source, regardless of who or what wrote the article. If the AI didn't make mistakes then it would be reliable. A human making lots of factual errors would also be an unreliable source. 

While of course humans can write shitty articles as well, AI as it is right now is pretty much incapable of journalism because it can't contact sources or witness an event. It's one thing to summarize your research and have the AI convert it to article form, it's another to expect it to pump out an article given just a title or a recycled press release...

[whispous]

AI generated media is slop. And it always will be.

AI doesn't know anything, and when it comes to content masquerading as any form of art, it's not art because a computer has no intent or message behind what it spews out.

 

SLOP.

[LAwLz]

3 hours ago, Sauron said:

While of course humans can write shitty articles as well, AI as it is right now is pretty much incapable of journalism because it can't contact sources or witness an event. It's one thing to summarize your research and have the AI convert it to article form, it's another to expect it to pump out an article given just a title or a recycled press release...

A large amount of articles written by humans do not contact sources or witness events themselves. They are just recycling press events or regurgitating what others have written. I am willing to bet that most of the news you are exposed to, at least if you are an average person, is not based on first-hand sources or what the writer was told by the original source in a one-to-one exchange.

 

My point is that the issue isn't necessarily AI, and this isn't a new issue at all. The issue is publications prioritizing pumping out articles with little to no fact-checking because "more articles/videos = more money".

I don't really see AI as being an issue at all. It's just that AI now allows these same publications to do more of the bad things they have always done. 

 

Remember, these AI tools aren't being operated by themselves. It's still a human feeding it what to write, and deciding to publish it. 

AI-written articles are not the issue here. If anything, AI-written articles are a symptom of the real issue, which is prioritizing quantity over quality and not fact-checking what they publish.

[OhioYJ]

21 hours ago, LAwLz said:

I think it's also worth mentioning that humans can be really shitty journalists too. 

I think many are these days. What happened to the five "W"s of journalism? I can't tell you how many news articles I read and still have no idea what happened? I should have stopped at the headline, as it was just as informative as the entire article. 

[Sauron]

12 hours ago, LAwLz said:

A large amount of articles written by humans do not contact sources or witness events themselves. They are just recycling press events or regurgitating what others have written. I am willing to bet that most of the news you are exposed to, at least if you are an average person, is not based on first-hand sources or what the writer was told by the original source in a one-to-one exchange.

I try to read as close to direct reporting as possible (or at least outlets that are clear about what sources they're referencing), but yes, a lot of online news is like that, which imo speaks more to the generally poor state of reporting rather than to the quality of AI articles. I did some work for a Linux blog once but I quickly gave up precisely for this reason; you were pretty much expected to pump out at least an "article" a day and there was no differentiation in pay or recognition between regurgitating other outlets and actually doing your own testing and research, which obviously took considerably more time and effort on my part.

 

What I'm trying to say here is that while humans can potentially do actual reporting, as rare as that is, AI pretty much can't, and therefore an outlet using AI en masse will never be able to be more than a shitty content farm unless the technology changes significantly.

[williamcll]

On 3/4/2024 at 2:04 AM, starsmine said:

Ill bite.
Why? 
Do they not strictly follow editorial standards and regularly correct themselves?

They should not be used as a source because of circular sourcing, not because wiki is wrong

It's supposed to follow editorial standards but different languages of Wikipedia (especially Chinese and Japanese) have different standards, especially reliable sources for politically related articles.