National newspaper duped into running GPT-4-written rage-click opinion piece The Irish Times, one of the republic's respected broadsheet newspapers, has apologized for publishing an opinion article said to have been generated by AI software and submitted by an anonymous trickster. Last Thursday, the publication ran the comment piece under the byline Adriana Acosta-Cortez, described as a 29-year-old …
When pushed to identify itself, GPT4 said:
"I do not have a name, I am a large language model designed to assist you".
We took to the streets and asked the public what they thought:
"Clearly, GPT4 is a mechanised version of Jacob Rees-Mogg with a personality sideloaded in, a boring one, but it has one, I won't vote for GPT4 at the next election".
Remember when the mainstream media used to try to work in a Titter angle to any minor story even when Twitter was only peripherally connected? AI is the new twitter.
Whether the story and byline photo were AI generated is utterly irrelevant to the main meat of this story. The primary issue here is that a major news outlet ran a piece that was utter tripe simply because they thought it would generate controversy. They didn't care a jot where it came from, they just saw it as something that would generate traffic and publicity in equal part.
The "fact" that it was AI generated is quite convenient for them. So convenient in fact that any reasonable person might wonder if the announcement it was AI generated was part of the plan. "Look lads, if we get called out on running this shoite what we do is claim we've been duped by a sophisticated AI scam. If that doesn't take the heat off we'll put Elon Musk in frame."
It was probably a better article than the stuff my 'Local' Reach PLC rag puts out. They write articles about disagreements people have on Mumsnet, and where to get free takeaway food. I only know this because I use an Ad blocker, the site is pretty hard to navigate without one.
Our local example is awful. It's obsessed with minor tittle-tattle, people being nicked for speeding, alleged pa*dophiles, and endless coverage of take-aways with 1-star hygiene ratings. I dunno what's worse, the mind-numbing nature of its triviality or the knuckle-draggers who comment on every one of those articles. As much as I may sometimes get irritated by some El Reg's esteemed Commentards™, the local rag's "discussions" really are down in the depths of the sewer. Oh, and even with ad-blockers it still manages to be irritating to read with its novel way of breaking scrolling. Mostly I don't bother unless something local is happening, and even then there's rarely anything of interest to be learnt anyway.
Our local rag reports every minor car bump that they hear about, sometimes with a picture of a cracked number plate or a car in a hedge.
The local pearl clutchers write in asking what's going on on the terrifying roads and the carmaggedon in our town. .
The technology [generative AI] has rapidly advanced, and is now capable of producing coherent-though-bland text and realistic images.
Yes, maybe also, but the expanding explosive existential problem for humanity on Earth, is that the rapidly advancing and quite alien technology is surprisingly capable of producing extremely rich and disruptively creative text for realistic augmented virtualisation/televisual representation/metadata physical production.
And whereas the possibility and sad likelihood exists that humans might dismiss and even propose some sort of vainglorious organised resistance against such a development/intervention, AI in its many phorms would have no doubt about its current abilities with present facilities for future utilities, and its almighty primacy in brave hearts and open minds and deep thoughts capture.
Over the years, I've gone back and forth on whether you were an account powered by AI, left to run rampant on El Reg's servers.
How does it feel to be less coherent than an actual AI?*
When are you upgrading to become amanfromMars 2?
*The person about to comment on ChatGPT etc not being an AI, we get it. You're very smart. Now be quiet - Mummy and the machine are sassing each other.
+1 for the (proper) use of "vainglorious".
Which, IMHO, goes to prove that amanfromMars1 is not a bot. No (current) LLM would create anything using that word, if for no other reason that the garbage it is trained on would have that word in it anywhere. (Well, yes of course, since LLMs will scrape the El Reg forums, it may show up somewhere in ChatGPT 5...)
This shows that LLMs are capable of: It is an opinion piece, full of assertions, with little argument or logic. No need to fact check. Nothing original (the same point of view has been expressed elsewhere in the past).
Unfortunately this is what journalists do. The Reg should not be too complacent - some of the opinion pieces here are pretty poor too.
The LLM did manage to insert some relevant links which is more than most journos writing for national newspapers see capable of.
Bullshit. This kind of false equivalence is not helping.
Yes, the IT's pearl-clutching has a whiff of self-exculpatory hypocrisy about it. And I'm not sure what it means by "coordinated" deception in this case - coordinated with whom?
But to suggest this is the same as what the media do all the time is not only to fundamentally misunderstand how the media works, it's also throwing in the towel and handing over the whole world to the Trumps and Putins among us.
There are no legitimate independent news organisations in Ireland
RTE is effectively state sponsored, relying on license fee increases and 20% ad content to fund the 2+ million a year the top ten presenters get
The Irish Independent stable is the Irish equivalent of the Daily mail
And the Irish Times is so stuck on the fence about where it wants to land that it's happy to solicit unvalidated, vaguely attributable articles to avoid any kind of accountability.
The mighty haven't fallen, except in the sense that they've expired and been progressively replaced by the trivial.
First, there was the stone age, when culture was defined by the use of stone.
Then there was the bronze age, when culture was defined by the use of bronze.
Then there was the iron age, when culture was defined by the use of iron.
Now, we're in the bullshit age and culture is defined by the use of virtual bovine excrement.
If human generated journalism can't be distinguished from the output of a glorified statistical auto-complete engine, what does that say about the current standards of journalism?
Now, we're in the bullshit age and culture is defined by the use of virtual bovine excrement.If human generated journalism can't be distinguished from the output of a glorified statistical auto-complete engine, what does that say about the current standards of journalism? .... Mike 137
Methinks, Mike 137, with specific regard to the above two sentences, one is going to have to accept as indisputable fact, many fundamental things have moved on a great deal further and a lot deeper and darker and higher and brighter than that, and they continue to race on ahead well clear of any and all competition or opposition to previously unknown, formerly unimagined spaces and really strange and spooky alien places, .... for the following is neither inaccurate nor fictitious ‽ .
Then there was the bullshit age when culture was defined by the use of virtual bovine excrement which humans couldn't distinguish from the output of glorious statistical auto-complete LLM engines and AI. ..... amfM
Now where are you? And do you have any almighty command and control there, in what may very well be is a virtually remote existential IT space with postmodern multi-media created places on Earth for live animal forms to inhabit and procreate in?
And is that last question, rhetorical, thus advising one and all that they are fully entered into a fait accompli situation in which they are more definitely passenger and spectator than driver and leader?
"while an AI system may be useful in suggesting information to include, don't trust it: check it against an authoritative source like you would any claim."
This. A thousand times, this. chatGPT et al are, indeed, useful tools. But they are ONLY tools. The sooner people get to grips with that, the better.
I used to be a regular contributor to the Irish Time's letters page.
At no point did they ever:
1. Verify that I existed and I was a real person.
2. Check that any content that appeared to be from me, was in fact from me.
I also have a friend who used to make a point of sending in provocative letters under false names. Even when the address he gave was that of a well known bakery within walking distance of the IT's office he was never challenged.
So the idea that they could be fooled like this doesn't surprise me. Another thing you have to bear in mind is that all media outlets have an agenda, even if it's an unconscious one, and content that aligns with that agenda will face less scrutiny than it should. In short, they are just as vulnerable to clickbait as humans are.
I was once told that half the letters in local rags and half the callers on radio call ins were from bored students. I was then later told that while they may be bored these students are actually doing it as part of their course work. Presumably you get credits for winding up Jeremy Vine.
I'm guessing contributors of such quality clickbait content get no freelancer / external contributor initial pay or syndication fees, it's just pure profit for the Irish Times?
If that's the case they have no incentive to verify as no money changing hands & "needless" verification would eat into their profits.
If payment was involved would expect some form of verification to be required.
We are safe, I must hope, from an editorial piece by amanfromMars? ....... Bitsminer
If one were to realise nowhere and nothing is safe and secure from anything anywhere, and such is a prime premium default for employment and enjoyment easily leading to rewards and benefits generated by the recognition that such a simple notion provides one with a much greater intelligence, or a different intelligence altogether, better servering one and all with vast hordes of inclusive mutually beneficial, positively reinforcing wishes, would the world/worlds be a greater place/different space?
Considering how chewy their posts are, at no longer than about a paragraph, a whole article written in that style would probably read with the undeniable grace and finesse of falling down several flights of stairs.
I'd genuinely love to see it. Trying to read it would be like exploring the depths of the Amazon Jungle, cutting through undergrowth while struggling to progress at all in the slightest. Knowing that every step risks your body & mind with the myriad ways your trek could go wrong!
One of the few remaining markets for science fiction (I forget which) is SOOOO overwhelmed with Artificial Idiocy, over 10X the usual submissions, that they stopped taking ANY submissions.
(True, sci-fi magazines often have excess inventory, locking-up good stories now and not paying until publication.)
They (and all and dog) say there is no decent way to filter this flood.
My sense from minor reading is that 90% of AI can be suspected after one or more minutes of reading. (AI spans the range of human 'thought' better AND worse.) Since slush-pile was already unmanageable, a 1,000% increase is 100% useless.
There was always a place for Writer's Agents, who know their authors/clients, and can testify "Human". But the agent business has been harsh for over a decade.
And why should we discriminate? Ron Goulart's novel Wildsmith, 1972, "Considered by many (who?) to be Goulart's best", is about a best-selling robot author, and his agent trying to hide that fact from his public.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
