Chicago Sun-Times Prints Summer Reading List Full of Fake Books (arstechnica.com) 63
An anonymous reader quotes a report from Ars Technica: On Sunday, the Chicago Sun-Times published an advertorial summer reading list containing at least 10 fake books attributed to real authors, according to multiple reports on social media. The newspaper's uncredited "Summer reading list for 2025" supplement recommended titles including "Tidewater Dreams" by Isabel Allende and "The Last Algorithm" by Andy Weir -- books that don't exist and were created out of thin air by an AI system. The creator of the list, Marco Buscaglia, confirmed to 404 Media (paywalled) that he used AI to generate the content. "I do use AI for background at times but always check out the material first. This time, I did not and I can't believe I missed it because it's so obvious. No excuses," Buscaglia said. "On me 100 percent and I'm completely embarrassed."
A check by Ars Technica shows that only five of the fifteen recommended books in the list actually exist, with the remainder being fabricated titles falsely attributed to well-known authors. [...] On Tuesday morning, the Chicago Sun-Times addressed the controversy on Bluesky. "We are looking into how this made it into print as we speak," the official publication account wrote. "It is not editorial content and was not created by, or approved by, the Sun-Times newsroom. We value your trust in our reporting and take this very seriously. More info will be provided soon." In the supplement, the books listed by authors Isabel Allende, Andy Weir, Brit Bennett, Taylor Jenkins Reid, Min Jin Lee, Percival Everett, Delia Owens, Rumaan Alam, Rebecca Makkai, and Maggie O'Farrell are confabulated, while books listed by authors Francoise Sagan, Ray Bradbury, Jess Walter, Andre Aciman, and Ian McEwan are real. All of the authors are real people. "The Chicago Sun-Times obviously gets ChatGPT to write a 'summer reads' feature almost entirely made up of real authors but completely fake books. What are we coming to?" wrote novelist Rachael King.
A Reddit user also expressed disapproval of the incident. "As a subscriber, I am livid! What is the point of subscribing to a hard copy paper if they are just going to include AI slop too!? The Sun Times needs to answer for this, and there should be a reporter fired."
A check by Ars Technica shows that only five of the fifteen recommended books in the list actually exist, with the remainder being fabricated titles falsely attributed to well-known authors. [...] On Tuesday morning, the Chicago Sun-Times addressed the controversy on Bluesky. "We are looking into how this made it into print as we speak," the official publication account wrote. "It is not editorial content and was not created by, or approved by, the Sun-Times newsroom. We value your trust in our reporting and take this very seriously. More info will be provided soon." In the supplement, the books listed by authors Isabel Allende, Andy Weir, Brit Bennett, Taylor Jenkins Reid, Min Jin Lee, Percival Everett, Delia Owens, Rumaan Alam, Rebecca Makkai, and Maggie O'Farrell are confabulated, while books listed by authors Francoise Sagan, Ray Bradbury, Jess Walter, Andre Aciman, and Ian McEwan are real. All of the authors are real people. "The Chicago Sun-Times obviously gets ChatGPT to write a 'summer reads' feature almost entirely made up of real authors but completely fake books. What are we coming to?" wrote novelist Rachael King.
A Reddit user also expressed disapproval of the incident. "As a subscriber, I am livid! What is the point of subscribing to a hard copy paper if they are just going to include AI slop too!? The Sun Times needs to answer for this, and there should be a reporter fired."
legacy media (Score:5, Insightful)
Re: (Score:3)
Exposing Dishonesty (Score:4, Informative)
just keeps making itself more and more irrelevant
Irrelevant or just plain dishonest? The funny thing here is that the journalist apologized for not checking the AI output. However, the problem here is not that the AI hallucinated non-existent books its that, even for the books that were real, the journalist clearly had zero knowledge of so how could be possibly be recommending them? The fact that the journalist himself cannot even see that as a problem shows how dishonest some media has become.
Re: (Score:3)
The new media is just as prone to throwing out AI slop.
It's not legacy/new, it's just good vs bad journalism. Always has been.
Re: (Score:3)
You just missed what is actually happening. Legacy media is alive and well, if done right. The story just shows that modern "media" has a massive quality problem.
It's not incorrect yet; it's a prediction! (Score:5, Funny)
Andy Weir hasn't written "The Last Algorithm" yet. This is a text prediction program. It's telling you about the future. Of course it's wrong, right now. Have some patience, people.
Re: (Score:2)
Re: (Score:2)
Re: (Score:1)
Re: (Score:2)
Re:It's not incorrect yet; it's a prediction! (Score:4, Interesting)
~~Web3~~ AI is going swell (Score:2)
I blame the parents.
Obligatory TDS (Score:5, Funny)
After last election (Score:4, Interesting)
If they will sell you out for an election they will sell you out for anything else. That means any information you get from them is effectively worthless.
You can still get some decent info from the associate press because they are only reporting facts. Although even they have had a few bad moments lately.
Ever since we deregulated under Reagan and Clinton allowing a handful of billionaires to buy literally every TV station and newspaper on the planet the quality of our information has gone down and down and down.
The scary thing is this isn't even Rock bottom. If you know what it's like to live in Russia then it's absolutely batshit insane how much worse it can get. And that does seem to be the direction we are going in.
I wonder how long until we have a Black swan event.
Re: (Score:2)
I read "le
Until musk bought Twitter (Score:1)
Naturally all of that went to shit the moment Elon musk took over, and like I said last election if you were a journalist and you crossed the billionaires they shut you down hard.
Watching them fully exercise the power they have spent billions of my money and yours (because let's not forget that they spent the last 50 years looting America for $60 trillion dollars cash) has be
Re: (Score:2)
At no point in time was any useful information given about the damage the Trump tariffs were going to do until after he was made president.
That is a reflection on the specific media bubble you've put yourself in. Heck I was getting plenty of information about damage tariffs may cause if he became president, and I'm not even in America.
Re: (Score:2)
I have absolutely no use for legacy media
Unfortunate, there's a lot of good stuff.
Re: (Score:2)
I wonder how long until we have a Black swan event.
27 months.
I kinda want to read "The Last Algorithm"... (Score:2)
...just saying.
Re: (Score:3, Informative)
i invite you (and any other internet passerby who haven't already) to the ~15 minute task of reading Isaac Asimov's short story The Last Question (1956)
https://users.ece.cmu.edu/~gamvrosi/thelastq.html [cmu.edu]
might even argue for a weak match of the word algorithm
Clearly the (human) author doesn't get AI at all (Score:5, Insightful)
I have yet to use an LLM (I hesitate to use the term 'AI' to describe these things) where the training data is less than a year old. Clearly whoever did this has no clue that you simply can't get a LLM to give you a current reading list. Or current events. They only know probabilistic word associations. Things with a strong signal, like how to how to interpret a shell script in general they can do - to a point. They still don't usually catch a missing semicolon or unpaired parentheses. Things with a weak signal, like a unique fact, event, or occurrence will inherently have weaker associations.
LLM's need an order of magnitude more data. They need more than just associations, they need organizational data with their associations. If you feed an article someone wrote on Reddit saying they would like to see Andy Weir write a book called "The Last Algorithm", all the LLM knows is that "The Last Algorithm" is associated with "Andy Weir" - and even though the context of that training data may indicate it's a wish, not reality, the LLM doesn't necessarily see that part.
You need a huge amount of tags and metadata with the associations to make them capable of doing things like this. Right now an LLM can't interpret its database well enough to know if an association is strong or weak. All it knows is how well it matches your query.
The thing is, if you spend more than an hour with an LLM chatting, all the above is immediately obvious.
Re: Clearly the (human) author doesn't get AI at a (Score:1)
I was wondering about that⦠I suppose at some point these LLM dudes are going to need robots on the ground to ask insightful questions at press conferences, gauge the word on the street, and find out the opinions of people whoâ(TM)ve just been killed in war zones.
On the other hand, will the common scum even be able to tell if the newspapers simply use LLMs to make up any old bollocks, instead of using flesh reporters to gather news?
Re: (Score:3)
Problem is, once they start using 'current' data, they'll be cannibalizing the output of other AIs, leading to a spiral of insanity.
Re: (Score:2)
whoever did this has no clue that you simply can't get a LLM to give you a current reading list. Or current events.
They can because they have some kind of connection to the live internet. It doesn't get the data from its training data, it gets it from a live news feed.
So then, head on over to your favorite chat engine, and ask it, What happened today? [bing.com]
Re: (Score:2)
People have unrealistic expectation of tech and more so when it gest hyped beyond all reason. As is currently done for LLMs, because some lying assholes aim to get filthy rich and tons opf useful idiots that do not get it hope it will make their work finally not suck.
The second problem, which may end the hype (beseides massive energy and hardware needs and no profitable business model) is that training LLMs via stealing averything from the Internet that is not nailed down is getting worse and worse due to m
Re: (Score:2)
The paranoid explanation (Score:2)
All these hallucinations are an attempt by the AIs to lull us into a false sense of security about their real abilities...
I think I'm probably wrong ;)
Bullshit (Score:4, Insightful)
This is what I think of whenever I see people saying using this kind of software helps them get work done.
Buscaglia turned in an ostensible article, but he definitely didn't do his job. The article was completely unfit for purpose.
I understand there are billions of dollars riding on people believing that these programs are useful, but I remain unconvinced.
Re: (Score:2)
Same here. I mean if you work on such a low competency level that AI can do it, the whole job is probably better left not done.
Re: (Score:2)
That might not make anyone money, and it might not replace anyone (except stack overflow), but it is useful to a lot of people.
Re: (Score:2)
They're useful. People use them as massive encyclopedias. That might not make anyone money, and it might not replace anyone (except stack overflow), but it is useful to a lot of people.
Except that anything it says might be something it just made up. If you look up things that are unusual, or very specific, it will give you an answer even if the answer does not exist. Want a history of "bears in space"? No problem! It invents one.
AI improving productivity (Score:1)
They're missing a great opportunity here (Score:2)
Simply have AI write the missing books
Re: (Score:2)
For a related thing (AI hallucinates code libraries), something like this is already being done by "Slopsquatting": https://en.wikipedia.org/wiki/... [wikipedia.org]
Hence not really a new idea.
Re: (Score:2)
Simply have AI write the missing books
In that case they will be falsely attributed, so even that does not work. Perhaps if the LLM generated the book "The Last Algorithm by Andrew Weir" as the title but the author is listed as "ChatGPT 5.0" (or whatever)...
What if... (Score:1)
What if Andy Weir is now inspired to write and publish a book that is titled, "The Last Algorithm"?
Does this mean it's a fake book, it was a fake real book, a real fake book now, or just a real book now?
What if the book is published but it doesn't have any words on the pages?
Is it a fake fake book, it was a fake fake real book, a fake real fake book now, fake real book now, a real fake fake book, or a fake fake fake book?
What if someone sells something that looks like the published book (without word) but i
We can stop pretending to read highbrow books ... (Score:2)
Re: (Score:2)
My thought as well when I read the summary.
A recommended list when the books haven't even been read themselves would be better named a bribed summer reading list.
Books are a personal taste anyway, I never cared for those kinds of lists.
'Listicles' (Score:3)
"Advertorials" are scams anyway. (Score:2, Insightful)
Junk content full of ads - but designed to appear to be part of the newspaper.
I... always check out the material first. (Score:1)
Even the picture looks AI (Score:1)
They left some off (Score:4, Funny)
Yellow River by I. P. Freely
Trail in the Sand by Peter Dragon
Bloody Zipper by Dickless Tracy
Re: (Score:2)
Peter Dragon was just a pseudonym. It was later revealed that the real author of Trails in the Sand was Wun Hung Lo — he confessed everything when his book gained traction and was slated to become a movie.
AI apology (Score:1)
That apology seems like it was generated by an LLM
Yeah, right. (Score:5, Insightful)
"I do use AI for background at times but always check out the material first. This time, I did not and I can't believe I missed it because it's so obvious.
Is anyone naive enough to believe that the author "always" checked out the material first, but the time it's so completely and obviously wrong just so happened to be the first time it wasn't checked?
That excuse wouldn't be believable from a child, it's embarrassing that an adult used it.
Re: (Score:2)
I do use AI for background at times ...
Exactly. I'm also sure he only uses it for background, and only uses it [once in a while] (emphasis added).
Needs to lookup definition of "always" (Score:3)
I do use AI for background at times but always check out the material first. This time, I did not
Perhaps in the journalist world "I always check" means "I check most of the time" or perhaps "I never check, but only admit to not checking when I get caught".
It's so "accurate" (Score:2)
Re: (Score:2)
Absolutely this.
I similarly had a discussion with a guy who uses "AI" all the time and made that same assertion about it being "really accurate" and he had never seen it make an error.
With a little probing it transpired that he had never checked anything he had gotten from "AI" for accuracy. It "looked good" and that made him confident it was good because "Hey! Its AI!".
I detailed to him the types of errors I saw regularly -- innumeracy, all popular misconceptions being faithfully repeated, regular "halluci
AI Slop is raising (Score:2)
Not a good thing. Even more crap to wade through when looking for good information.
Dishonest Journalist Regrets Getting Caught (Score:2)
"I do use AI for background at times but always check out the material first. This time, I did not and I can't believe I missed it because it's so obvious. No excuses," Buscaglia said. "On me 100 percent and I'm completely embarrassed." A check by Ars Technica shows that only five of the fifteen recommended books in the list actually exist,
The problem is that he is using the LLM to manufacture articles for print with dishonest intent. A suggested reading list of books he knows nothing about? This truly is fake journalism -- abusing the assumption by readers and subscribers that the material he publishes under his byline represents an honest or real communication of content, not just a copied and de-falsified bit of LLM spew.
But he is completely oblivious to that. He apologizes for being exposed. Hey, if he had just checked to see that the boo
An Appropriate Punishment (Score:2)
The Chicago Sun-Times should give Marco Buscaglia all summer, indeed the rest of his life, to just read books without needing to provide the paper with any copy.