Brit newspaper giant fills space with AI-assisted articles Reach, the owner of the UK's Daily Mirror and Daily Express tabloids among other newspapers, has started publishing articles with the help of AI software on one of its regional websites as it scrambles to cut costs amid slipping advertising revenues. Three stories written with the help of machine-learning tools were published …
Nope.
Writing is best unless it's something you specially have to see to understand.
Why? Because I can skip down the headings to the part of interest.
With video, there's the intro, then some content, then begging for subscribers, then some content, then something about Patreon, then some content that sounds suspiciously like something that had already been said, then it segues right into a sales pitch for the sponsor (that's probably something completely irrelevant like meals-in-a-box). And then "don't forget to subscribe, bye" and, wait, what?!
Grrr!
Give me words any day.
We're still doing a lot of written word, as you can see, and some videos, which are fun, useful, and a break from writing.
There's a reason why steakhouses have some chicken and pork on the menu as well as beef. Little bit for everyone. Nothing wrong with a bit of variety for those who prefer listening or watching.
Our videos aren't like YT where people beg for Patreon and likes and subscribes. For us, it's The Register logo, and then right into the chat, and we keep it short. It works as audio-only too if you want to listen while doing something else.
C.
The majority of newspapers stopped reporting on the News a long time ago. The editors allowed the reporters to spew their ideological political views instead of sticking to the facts and allowing the readers to make up their own minds.
The end result was that the vast majority of the public, in all Western countries have nothing but contempt for the press.
Hearing that these same 'journalists' are all losing their jobs gives me great satisfaction. It really does. Now they are free to spew their sanctimonious hypocritical viewpoints on twitter, where I can safely ignore them.
Couldn't have happened to a nicer bunch
No, "opinions" spewed as facts are what gave us Fox, Newsmax, OAN, Trump, Jan 6th etc. and they are the reason the country is in such a mess.
On the other side, you get MSNBC posting pro-corporate-democrat stuff, but at least they aren't out and out liars like the right-wing swamp.
Any chance of journalists actually posting the FACTS, without adding an agenda, or being afraid of what their advertisers, or politicians think?
None whatever.
Nobody, literally nobody including you, is interested in "facts". They want "stories", which is what reporters have always written.
Asking reporters to post "facts" is like asking builders to deliver a pile of bricks, or restaurants to bring you a heap of raw ingredients - but without specifying anything about what you're trying to build or make. And then people will get more upset because the windows are the wrong shape or the ingredients don't include any meat, or something.
There's a good Tom Scott video talking about truth and facts which is what's being discussed here. (It is a bit long, but worth it)
Obviously you don't know what "literally" means. Unless you live in a bubble chamber of insecurity, you are interested in facts.
Now, I'm sorry I didn't write "the facts should be presented within sentences, in a logical order, showing additional context or background along the way", because I assumed that was a given - I wasn't saying news show should just consist of monotonic reading of lists from a script, just that it shouldn't be a work of fiction.
So, pretty crappy strawman. You know full well what I meant when I said "facts without a personal agenda'.
If you need to hear stories, quit following politics, or news, and go immerse yourself in soap operas. It's all you seem capable of understanding.
The M-E-N is more like someones twitter feed than a news website.
(ad) (ad) (ad) (ad)
Sensational headline (We're all going to starve!!!), two paragraphs of fluff, then one sentence saying that the price of baked beans has gone up in a local shop.
(ad)
Half-a dozen pictures from the archives (now spread across as many pages to generate more clicks).
(ad) (ad) (ad)
Latest celebrity promo.
(ad) (ad) (ad) (ad) (ad) (ad) (ad) (ad) (ad) (ad) (ad) (ad) (ad) (ad) (ad) (ad) (ad) (ad) ...................................................
I like Kevin Bloody Wilson's view of all this crap ... the point is that YouTube has no adverts shown in most of his videos!
Looking at my local papers online presence, hovering over one advert link shows as something like https//[loads of letters].com/[loads of letters]
but when I right click to copy the link, it turns out to be https//paid.outbrain.com/network/redir?p=[loads of letters]
Bad coding, or trying to obfuscate?
/redir?p=[loads of letters]
Your page request goes to the tracking site, they log all your details, then they redirect you to the site you wanted. They also get to control what the target site knows about you. Simplest anti-tracking workaround is to just search (in a new tab) for the subject of the link, which should show the site you wanted near the top. But don't search using Google, because their result links also use the redirect tracking technique.
My local rag, the Lester Mercroh is a Reach publication. The 'news' articles are often 'journalists' telling us some people on Mumsnet had an opinion about something, where to get free fast food, or that people watching some TV program were upset. I only know this because I use an Ad blocker, otherwise the site is unreadable.
AI written articles can't be worse, surely.
Those ads that jump around while you're trying to scroll to find the content. Then the giant notice which says "We respect your privacy"... By using 600 tracking scripts on the page to max out the ad revenue.
The advertising revenue will drop if they're all from accidental clicks and 80% of the screen real estate being stolen by it. Very bad experience for readers. But they are good with their clickbait titles.
That's not just news websites, though. That's pretty much endemic to the internet as a whole.
The reasons for which were laid way back when "HTTP" was first developed, and the idealists at the time decided the internet should be free to use. Seemed like a great idea at the time. Now we can see the downside.
ChatGPT content at Retch Media will actually be a significant improvement from the normal hysterical Beast from the East, new McDonalds/Tim Hortons opening, advertiriala masauerading as news and national news stories piped into local media they own … like the formerly respected Coventry Telegraph..
The spelling and grammar should improve too.
no idea why your post got a download, I LOLd at 'Retch Media'.
I presume the fast food articles are advertorials really, and sponsored, as they certainly aren't news.
On spelling and grammar,... this article said they'd proof read AI created content,... well, plenty of spelling and grammatical errors slip through the meatsack generated content, so I think few articles get checked as it is, so I doubt many AI articles would in the end, they'll just publish and be damned.
"Write an article about a new branch of McDonalds opening in Poundbury, Dorset."
Here is a possible article based on your request:
McDonald’s opens new branch in Poundbury
McDonald’s, the world’s largest fast food chain, has opened a new branch in Poundbury, an experimental urban development on the outskirts of Dorchester, Dorset. The branch is located on Queen Mother Square, a central plaza named after the late Queen Elizabeth The Queen Mother.
The new branch is part of McDonald’s expansion plan in the UK, where it aims to open 200 more restaurants by 2024. The company said it chose Poundbury for its unique design and vision, which is based on traditional architecture and New Urbanist principles as advocated by King Charles III when he was Prince of Wales and Duke of Cornwall.
“We are delighted to be part of Poundbury’s vibrant community and offer our customers a convenient and enjoyable dining experience,” said John Smith, the manager of the new branch. “We have designed our restaurant to blend in with Poundbury’s distinctive style and character, while also providing modern amenities and services such as free Wi-Fi, self-order kiosks and table service.”
The new branch has created 40 jobs for local residents and will operate from 6am to 11pm daily. It will also participate in McDonald’s environmental initiatives such as recycling packaging waste, reducing plastic use and sourcing ingredients from local suppliers.
Poundbury is home to more than 4,600 people and 240 businesses, including shops, cafés, a garden centre and an art gallery. Visitors to Poundbury can use free parking to explore its attractions or enjoy its events such as farmers’ markets, craft fairs and concerts.
McDonald’s is one of the latest additions to Poundbury’s diverse food scene, which also includes Italian restaurants, Indian takeaways, bakeries and pubs. The new branch hopes to attract both locals and tourists who are looking for a quick bite or a family meal.
From today, no less. Eating fatty foods linked to vision loss later in life... https://www.express.co.uk/life-style/health/1743468/vision-loss-risk-factor-high-fat-diet
(no, I'm not making that a clickable link, I have standards, and only noticed that article because it turned up in my news app)
There is no intelligence in these AI systems.
They have merely automated the process of googling a few articles and cutting and pasting the results.
With no real journalists to write the original pieces the whole thing falls apart. We just get a bunch of applications blindly copying each other and producing nonsense.
Although given the current news output it may be a while before anyone notices.
We just get a bunch of applications blindly copying each other and producing nonsense"
Even when not nonsense, the duplication is far from helpful. I just attempted to find an informative review of a non-fiction book via five separate search engines/proxies. Not a single independent review -- just two alternative copycat blurbs that look like verbatim rip offs from the jacket and the publisher's promo.
This is no longer the information age.
This post has been deleted by its author
With no real journalists to write the original pieces the whole thing falls apart. We just get a bunch of applications blindly copying each other and producing nonsense.
Have you read the average local rag? Mine is 74% undigested press release, 24% "underpaid junior reporter gets a paid jolly and says paying corporation is wonderful" and 2% "Snow/Ice/Beast From the East/Ice age Glaciers to return next week!!!".
Certainly in the UK redtops, "news" is already an algorithmic process. Generally based on reflecting the views of the readership back to them with words like "EXPLODE", "FURY", "UNLEASH" as separators.
Anyone who has the MS Start feature enabled will know this - and endless parade of clickbaity stories that all push an agenda.
I was immediately reminded of this: https://www.theregister.com/2008/06/10/have_your_say/. Unfortunately it looks like the twat-o-tron has gone 404 - perhaps succumbing to a stroke in a fit of virtual apoplexy, but more likely has been redeployed to writing policy documents for the home office.
I leave several in my books. It helps when some [redacted][redacted][redacted] numpty steals the text and publishes it on Amazon under their name despite the original also being on Amazon.
AMZN can be real dickheads about taking down the plagiarized version even with a DMCA notice and proof that it was stolen.
Now that ChatGPT and other pretend AI systems... Show me their Turing Test results please... are being used as an excuse to lay off journo's we are getting to the point where it won't be long before even the TV news readers will be deep fakes. We truly are doomed.
The street where I grew up was one of those trap streets. Actually it was an access point for the utility company, but a well-known map maker drew it as a through way. Their anti-plagiarism device failed abjectly, because eventually *all* maps showed it as a through way. The town placed huge boulders across the end of the street to prevent numpties from trying to drive through. Sometimes desperate numpties would find their way around by driving through somebody's side yard, only to find more boulders on the other side blocking their way out.
When all content is plagiarised from somewhere, how can plagiarism laws be enforced? "Your honor, it's true we plagiarised, but plaintiff doesn't have standing because we plagiarised it from someone else."
OpenStreetMap.org refers to these (non-existent roads or deliberate spelling errors) as Copyright Easter Eggs: https://wiki.openstreetmap.org/wiki/Copyright_Easter_Eggs
For ages, Bing had a town near me named "Empshott", presumably a deliberate misspelling of Kempshott which isn't a town and wasn't quite where they said it was.
Reach "local" news sites I have visited are a waste of space (& adblockers a must as without they take an eternity to load & even then sometimes crash randomly) as there's little in the way of actual news, just plenty of click bait style headlines.
If they do have some local news it's typically published far later than on the BBC local area section
> If they do have some local news it's typically published far later than on the BBC local area section
Well, duh - they have to retype it, that takes time (thank heaven for the CapsLock key or they'd learn how to use *two* fingers at once, to type Shift+Letter, and then it is only one step to Ctrl-C on the BBC website).
This is a first, their usual output is so dire that even so-called AI must be an improvement. I struggle to see how AI is going to cover the topic occupying the space not taken up by Love Island and Katie Price though - namely what the local free food they tried was like. I suppose the AI can review Spam though….
I usually conclude that whoever wrote a Reach article failed all their GCSEs, and did so without even starting their English one. So an AI might actually be an improvement to grammar and hints of being coherent.
Let alone the lack of actual content, the clickbait headlines and the million ads.
Reach already owns The Daily Star which has already reached rock bottom in terms of quality and started digging using heavy machinery. I'm far from convinced that adding AI's is going to further reduce the quality in this case; probably just the opposite.
To make the point; a number of years ago the Royal Navy decided to buy some new 5 inch guns for the new warships being built. These are the guns sitting on front of a big warship, which are 5" wide, and throw quite big shells.
To quote from elseware "When the contract was made public, British tabloid journalism reached a new level of hilarious incompetence, as the Daily Star newspaper published a piece claiming the MoD had “blown £183 million on a 5-inch gun which is the length of a toothbrush”.
link in case you think i'm making that up.
So yes, the quality output from Reach can hardly get worse with the addition of an AI; you'd hope that an AI would avoid these sort of errors. (Mind you, you'd hope that humans would as well...)
"a 5-inch gun which is the length of a toothbrush"
It's a grammatical error rather than a factual one. The modifying clause "which" refers to "5-inch" but is next to "gun". I can imagine some poor journalist trying to think of a grammatically correct way to say the same thing without also violating the style guide. Eventually -- fuck it, I know it's wrong but I don't care any more. Using an AI (sic) will be a big improvement, because it doesn't care right from the start.
Given the sorry state of some of the "stories" I've read on local news websites: these 'editors' need to be rounded up and forcefully educated with basic language skills (including 'reading') because the standard of English (let alone the standard of 'journalism') in these 'websites' (read 'advertising canvases') is truly dire.
Use an LLM to create the content [1]
Use another LLM to read the content and post comments on the "articles" (making use of LLM's ability to rant) [2].
Post "engagement figures": win awards for "audience figures" and, on the back of those, lots of lovely ads [3].
Rake in the bucks [5]
[1] At last, a proper use for "create" instead of "write".
[2] We have some especially trained on TOWIE, Made in Chelsea and Love Island to get a realistic(!) spread of reactions.
[3] Ad engagement? We have a Javascript macro for that [4]
[4] Don't worry about "being caught out" with fake ad reading, the so-called analysis of whether ads are actually worth the money is all codswallop: we can just write our own opinion pieces about how well ads perform and they'll lap it up
[5] With any luck, they'll realise all of this can be done within a single data centre, saving all the ISP bandwidth costs, and the dross will fade from public view. Wild optimism, I know.
Write me an article on the ethics of AI in they style and humour of the register.com
Artificial intelligence has come a long way since its inception. From chatbots that could barely string together a coherent sentence to advanced machine learning algorithms that can beat humans at complex games, AI has made some impressive strides. But with great power comes great responsibility, and there are some serious ethical questions that we need to address.
Let's start with the obvious: job displacement. As AI becomes more capable, there's a real risk that it will start replacing human workers in various industries. We're not just talking about low-skill jobs like factory work or data entry, either. AI is already being used to write news articles and generate financial reports, and it's only a matter of time before it starts encroaching on more advanced professions like law or medicine.
So, what do we do about this? Some argue that we need to implement a universal basic income to support those who are displaced by AI. Others say that we need to focus on retraining programs to help workers transition to new roles that can't be automated. And then there are those who argue that we need to slow down the development of AI altogether to prevent it from becoming too powerful.
But job displacement is just the tip of the iceberg when it comes to AI ethics. There are also concerns about bias and discrimination. AI algorithms are only as good as the data they're trained on, and if that data is biased, then the algorithm will be biased as well. This is already happening with facial recognition technology, which has been shown to have higher error rates for people with darker skin tones.
To make matters worse, AI is being used to make important decisions that affect people's lives, such as whether someone is approved for a loan or whether they're a good candidate for a job. If these algorithms are biased, then they could perpetuate discrimination and inequality.
So, what's the solution? For starters, we need to be more transparent about how AI algorithms work and what data they're using. We also need to make sure that diverse groups of people are involved in the development of these algorithms to prevent bias from creeping in. And we need to hold companies accountable for any harm caused by their algorithms.
AI is an incredibly powerful tool that has the potential to do a lot of good. But if we're not careful, it could also do a lot of harm. We need to take these ethical concerns seriously and make sure that we're using AI in a responsible and ethical manner.
How often are these "AI"s used by newspapers going to be retrained - and even then, how often are the training datasets going to be extended with text that is actually up to date (and have the by-then-demonstrated inaccuracies cut out)?
Even for cheap "listicles" like the one linked to in the article there is a point when the data gets old: thankfully, it hasn't happened, but what if Tredegar House had been closed for the last six months due to a fire, not expected to open for a couple of years? As it is, Tredegar is currently closed due to bad weather, so don't use that listicle as a good indication of what you can do this week whilst staying (mostly) indoors.
Do we think that Reach are bothered by this? Do they look bovvered?
[1] And this was reported in the knitting column, not even on the front page of the Women's Pullout Extra! Don't be surprised when the Actual News gets buried because it can not possibly be important: only one item in today's new training data has mentioned this so-called "Four Minute Warning", and the closest match to anything else indicates that this is clearly a new recipe idea for the perfect poached egg.
... whether Princess Diana is still dead?
The night she died I couldn't sleep due to a combination of hot humid weather and insomnia, so had the radio on for the World Service on Radio 4. As news of the accident came in it switched back to Radio 4 doing endless loops of talking heads not knowing much trying but desperately to fill airtime, with half hourly bulletins saying words to the effect of "Princess Diana has been seriously injured in a car crash in Paris, if there's any change in her condition we will let you know". Then as news of her death arrived they managed a mangled bulletin saying "Princess Diana has died in a car crash in Paris, if there's any change in her condition we will let you know" to which I thought "if there's a change in her condition, they should let the Pope know".
About a year ago I investigated as to why Reach were heavily invested in DWP marketing like articles.
Not being able to pinpoint the same trait down to ine person I deduced they were very low paid and on piece work.
Maybe it was always a chatbot like device.
With media far right, initially I suspected it was done in league with government. In a way it is.
The Express is another chatbot like place.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
