What's going on here?
-
What's going on here? The matplotlib maintainer this story is about correctly notes that all the quotes from his post in the article are made up.
UPDATE: Link was pulled; see below.
@mttaggart
I don't know what is real anymore, holy shit.If I'm reading things correctly Ars hallucinated the quotes in the article but still linked to this blog post correctly, right?
https://theshamblog.com/an-ai-agent-published-a-hit-piece-on-me/
-
3/
Koch Network
Building a White Supremacist Constitution one brick at a time over 60 years.The hate is now automated, corrupt, & well funded.
https://www.nytimes.com/2025/12/20/us/politics/koch-network-2024-election-trump.htmlhttps://www.desmog.com/2025/12/11/the-koch-network-is-pushing-trump-to-accelerate-ai-documents-show/
https://progressive.org/latest/like-dad-charles-koch-bircher-new-documents/
With AI, the wealthy skip paying for hate-for-hire campaigns like Facebook, Tenet Media, or the Russian Internet Research Agency
https://www.democracynow.org/2014/7/8/kochs_anti_civil_rights_roots
https://www.newyorker.com/news/news-desk/who-sponsored-the-hate
https://thehill.com/opinion/white-house/5326284-birchers-conspiracies-trump-legacy/
https://www.thebulwark.com/p/did-the-john-birch-society-win-in-the-end
@Npars01
Imagine having half a BILLION dollars to change the world, and choosing to bring this about! -
@dusk @cR0w @mttaggart @iampytest1
"…I don't know what we'll be able to say, so I don't want to over-promise anything…"
You say, 'We won't use AI to write our publication'
There's nothing else to say.
There's no promises to be made.Ars prides itself on technical accuracy. AI slop is a direct hit to their identity. The longer they wait to give a straight answer, the more it *appears* like they're trying to figure out how much they can get away with, instead of just being bluntly transparent.
@rusty__shackleford @dusk @cR0w @mttaggart @iampytest1
This is bad enough that a serious company that wanted to salvage their reputation properly might wanna consider putting in some weekend overtime.
But if they were being conscientious and diligent none of this would've happened in the first place. So of course instead it's, oops oh well I guess we'll deal with it sometime next week or something...
-
@mttaggart
I don't know what is real anymore, holy shit.If I'm reading things correctly Ars hallucinated the quotes in the article but still linked to this blog post correctly, right?
https://theshamblog.com/an-ai-agent-published-a-hit-piece-on-me/
Yep, that's about the size of it!
-
@theorangetheme @mttaggart oh, we very much don't, and I've even had editors insert mistakes into my stories before, without giving me readbacks. But this isn't the case of a single word or phrase, this is entire quotes being inserted. I don't buy into the idea that there is pressure on journos to use AI though — the point of the profession is original work, which AI by definition cannot do.
@aliide @theorangetheme @mttaggart
You would think that your point about original work would apply to software engineering too (I think it does), but there seemingly are managers out there - or at least directors applying pressure - who want their engineers to use it. -
@tankgrrl @mttaggart I mean, I assume that's what an internal investigation was about?
They probably want to properly call the author and ask them if they used AI or not, what were their sources, etc.
I don't think it's fair to mock them for wanting to conclude an investigation. -
@aliide @theorangetheme @mttaggart
You would think that your point about original work would apply to software engineering too (I think it does), but there seemingly are managers out there - or at least directors applying pressure - who want their engineers to use it.@GerardThornley @theorangetheme @mttaggart
I think the pressure on SWEs is *considerably* higher given the institutional pressure to make AI work! I'm sure many engineers want to write original code too, though.
-
@GerardThornley @theorangetheme @mttaggart
I think the pressure on SWEs is *considerably* higher given the institutional pressure to make AI work! I'm sure many engineers want to write original code too, though.
@GerardThornley @theorangetheme @mttaggart
A bigger problem with AI and journalism is probably the volume of writers who have written about it who don't seem to understand that the fearmongering about it is part of the hype. Like the "AI could take over the world" spiel is definitely perpetuated by companies and individuals that want it venerated as some out-of-control force beyond human comprehension. It's just another side of the AI hype coin.
-
@mttaggart@infosec.exchange link is dead
@addison @mttaggart It‘s in the Internet Archive.
-
@GerardThornley @theorangetheme @mttaggart
A bigger problem with AI and journalism is probably the volume of writers who have written about it who don't seem to understand that the fearmongering about it is part of the hype. Like the "AI could take over the world" spiel is definitely perpetuated by companies and individuals that want it venerated as some out-of-control force beyond human comprehension. It's just another side of the AI hype coin.
@aliide @theorangetheme @mttaggart
Yeah, I do get a little frustrated with the big "skynet" doom scenario stories that for the time being are unlikely, when things like obfuscation of responsibility, (at least attempted) manipulation of populations, and drastic economic shifts are pretty much here already and certain to cause harm. -
What's going on here? The matplotlib maintainer this story is about correctly notes that all the quotes from his post in the article are made up.
UPDATE: Link was pulled; see below.
@mttaggart The thick plottens.
-
Was AI used to generate this content? (Yes/No)
Does Ars have a strict "No-AI" policy for editorial content? (Yes/No)
If the answer to both is "Yes," how did the internal vetting process fail?
Regardless of a holiday, "I don't know what we'll be able to say" implies negotiation with the truth. For a publication built on facts, the only thing to "say" is the truth of what happened. The longer the silence, the more it looks like calibrating an excuse rather than admitting a failure.
-
Was AI used to generate this content? (Yes/No)
Does Ars have a strict "No-AI" policy for editorial content? (Yes/No)
If the answer to both is "Yes," how did the internal vetting process fail?
Regardless of a holiday, "I don't know what we'll be able to say" implies negotiation with the truth. For a publication built on facts, the only thing to "say" is the truth of what happened. The longer the silence, the more it looks like calibrating an excuse rather than admitting a failure.
@Gaelan
Strategic ambiguity is what this *appears* to be, it's the calculated vague speak that allows for plausible deniability that gets me.Also, news cycles: Friday news dumps allow stories to die over the weekend. Pushing the response back isn't just about the holiday, it’s about waiting for the news cycle. They're betting that by Tuesday, the "outrage" will have lost its momentum, making vague statements easier to swallow.
I know they have internal processes for this, but not a good look.
-
@mttaggart I suppose one thing you can say about old newspapers - no one can pretend an article was never printed
@Epic_Null @mttaggart
Winston Smith's job in 1984 was to change newspaper stories to match the Party's version of the truth, and the original sent down the Memory Hole to be incinerated. -
@colo_lee What anti-trans piece?
@iampytest1 It's fair to ask for a citation. I can give you some clues: it was shortly after musk acquired Twitter. Story by Timmer, I believe. More like unconscious and unacknowledged bigotry than the journalistic malpractice of this latest piece. Trust breaking behavior on the part of Ars editors and mgmt. I cancelled my sub at that point. I've seen a lot of links to Ars articles recently and thought about starting to read them again. But ...
-
What's going on here? The matplotlib maintainer this story is about correctly notes that all the quotes from his post in the article are made up.
UPDATE: Link was pulled; see below.
@mttaggart this is the weirdest story. Here is a link to SCOTT SHAMBAUGH’s blog explaining the whole thing with an update about the additional AI generated reporting. https://web.archive.org/web/20260214062635/https://theshamblog.com/an-ai-agent-published-a-hit-piece-on-me/
-
@aliide @theorangetheme @mttaggart
Yeah, I do get a little frustrated with the big "skynet" doom scenario stories that for the time being are unlikely, when things like obfuscation of responsibility, (at least attempted) manipulation of populations, and drastic economic shifts are pretty much here already and certain to cause harm.@GerardThornley @theorangetheme @mttaggart yes! as well as the problems/biases inherent in the training material or in the ways that it's trained
-
UPDATE: They pulled the story, but I had it up and had SingleFile in my browser, so: https://mttaggart.neocities.org/ars-whoopsie
@mttaggart Revenge of the bots. Creepy.
-
@mttaggart
"But that person has not come forward. If they instructed the agent to generate the blog post, they bear responsibility for a personal attack on a volunteer maintainer. "
Sounds like a solid case of #Libel to me@JizzelEtBass @mttaggart even if they *didn't* "instruct" the tool to do so, they're responsible for the text it generated, I'd say
-
@mttaggart this is the weirdest story. Here is a link to SCOTT SHAMBAUGH’s blog explaining the whole thing with an update about the additional AI generated reporting. https://web.archive.org/web/20260214062635/https://theshamblog.com/an-ai-agent-published-a-hit-piece-on-me/
I am so disappointed in Ars Technica.
