With a title like that you’re probably expecting a detailed, sophisticated analysis of every newspaper output and blog post in multiple languages worldwide. Unfortunately, no. It’s a deliberate homage to the dramatic titles used by everyday posters that end up either being about nothing or written by ChatGPT.
Almost every morning I read my curated Google News feed on my phone and invariably have to click on the three dots and choose “Hide all stories from_______” when another of my favourite news services has served up AI generated rubbish (slop) or in the case of those I gave extra chances to, either very long, or multiple AI articles (superslop). I was almost going to hide stories from The Conversation recently when I realised that I don’t see that much slop on there. Probably one in 20 articles that pop up for my chosen topics. So, The Conversation is staying on my feed for a bit longer.
Since at least 2010, blogs have been automating articles. It usually means Google buries them in the search results and specific articles are only found if they cover the particular thing a user is looking for, in which case they succeed. But generating revenue from ads and affiliate commissions that way doesn’t work for everyone. Now, with costs rising, even well-known brands are generating LLM articles, with em dashes everywhere. The problem with these articles, which may be the main subject of another post, is that information seekers have shifted from search engines to answer engines. Many of us just want an answer to a question, not sit and read an article for half an hour.
[Digression] What I’m looking forward to is when language is not a barrier and answer engines answer questions based on information from articles from any language from any country. That will be an exciting time for knowledge expansion. We’re not there yet, though. Most AEs mainly give us answers either from the USA or from our own country, but rarely anywhere else unless we specifically prompt them to. [Digression end]
With even major news publishers allowing LLMs to write articles, due to the need for advertising revenue and so constant publishing of content, these will only increase in quantity even as more readers move away from them – making daily news a superslop glut of staidly written misinformation.
While traffic to slop articles is dying, traffic to real articles aren’t as much. You may have already found that ache in your head from reading set news prose containing words at a set reading level in a repetitive order, miraculously disappears when you’re reading an interview, or when someone has written it from their point of view with a lot of ‘I’s, or used phrases that reflect the writing style or character of the publication or writer. I recently discovered this article in Popular Mechanics on the latest in Warp Drive development: https://www.popularmechanics.com/science/a65653811/scientists-say-physical-warp-drive-is-now-possible/ Have a quick skim of the link above before continuing, and see if you can see why this journalist-written article is so much more interesting to read than an LLM-written one. For comparison I asked ChatGPT to write its own version. I’ve italicised words and phrases in the ChatGPT version that, in my mind, causes a brain ache!
—
Warp Drives: From Science Fiction to Sub-Light Reality (by ChatGPT)
For decades, “warp drive” has been shorthand for faster-than-light travel—an icon of science fiction where ships slip the bonds of Einstein’s cosmic speed limit. But a new study by Alexey Bobrick and Gianni Martire suggests the concept might not be as far-fetched as once thought—if we’re willing to set aside the fantasy of instant galactic hops and work within the physics we know today.
The most famous theoretical design, Miguel Alcubierre’s 1994 warp drive, works by compressing space in front of a vessel and expanding it behind, letting the craft “surf” a moving bubble of spacetime. While clever, it has a deal-breaking flaw: it requires exotic matter with negative energy density—on the order of a star’s mass for even a meter-scale bubble. This isn’t just a lot of energy; it’s the kind of stuff our universe may not allow us to produce in bulk at all.
Bobrick and Martire’s research takes a different tack. Instead of clinging to Alcubierre’s constraints, they generalize what a warp drive can be. In their framework, a warp drive is essentially a moving shell of matter—positive or negative energy—that alters the properties of the space inside it. Crucially, they show that certain sub-light “warp bubbles” could be built entirely from ordinary matter, slowing time inside the bubble relative to the outside, without breaking any known laws of physics.
Their analysis classifies warp drives into four categories, from “mild subluminal” designs (realistic in principle) to “extreme superluminal” ones (still in the realm of speculation). For the realistic types, the bubble’s outer shell acts like a gravitational lens, subtly reshaping space and time for the passengers within. These drives won’t outrun light, but they could, in theory, carry people or payloads with unusual efficiency or unique timekeeping advantages.
They also explore engineering optimizations—like flattening the bubble shape to reduce energy demands by orders of magnitude. This tweaking of geometry might not get us to Alpha Centauri, but it could make experimental “warp-like” effects achievable in laboratory analogues, where light or sound waves mimic the behavior of spacetime.
The big takeaway? A warp drive doesn’t have to be a magic ticket to instant interstellar travel. By loosening Alcubierre’s rigid requirements, physicists can imagine—and even start testing—designs that operate with positive energy and sub-light speeds. That’s a far cry from the Enterprise (ChatGPT italicised), but it’s also a step toward turning warp drive from a purely fictional plot device into a legitimate branch of experimental physics.
In the words of the authors, “warp drives are simpler and much less mysterious objects than the broader literature has suggested.” The real mystery now is how far we can push them—and whether our first working warp bubble will fit in a lab rather than a starship hangar.
—
Can you see the difference? Caroline Delbert’s article is interesting to read, with expressions like “nailed down” and “flies in the face of” and “crazy concept” just in the first main paragraph. Using a unique example like “tablecloth and dishes” , including a video to demonstrate that (though that might have been an automatic link) using “seriously” to emphasise, “pretty thick” to describe, and the genuinely fun image of “like having Albert Einstein show up to your introductory physics class.” I also like there’s a specifically described link to a video on warp drive news, while useful in context, it subtly indicates that the journalist did some extra research besides just rewriting a science article. i.e. there is joy, originality, creativity, and a certain je ne sais quoi that just doesn’t exist in an LLM-written article. After all, one of the aspects of reading is to give us pleasure, to have that dopamine hit from enjoying someone else’s creation.
Looking at the LLM written version, (my italics), its just that little bit off – which I’ve just come up with as representing “the uncanny valley of writing.” (Now, I’m sure others might have come up with that already so I’ll have to do a search later.) But it perfectly explains what I’m talking about. Decades of reading and developing an expectation of what we’re going to see, even in a creative way, don’t quite hit the mark in a ChatGPT article, and some creativity doesn’t make sense. If a blog post writer like myself can easily correct these, why can’t ChatGPT?
Examples:
- Why “deal-breaking flaw” instead of “a deal-breaker”?
- How does the universe stop us? Isn’t it physics?
- Is power produced in bulk?
- It’s an article on theoretical physics. Why use ‘unusual’?
- What’s timekeeping got to do with it? “It’s all relative” would have been better.
- Why use ‘built’ with the bubble metaphor when bubbles are considered organic?
Oh, the pain. The pain.
Where’s that je ne sais quoi, ChatGPT?
So, no, articles aren’t dying. We just need to put in a bit more extra time to find the good ones.
If you’ve reached the brain ache limit on LLM generated articles, do what I do and read science articles from early newspapers. Check out Trove’s vast Australian newspaper archive here: Trove Newspaper Search
Update 19 November 2025 – While articles everywhere now look like they’ve at least been edited or translated by ChatGPT and other LLMs, if not completely written by them, a new trick has been implemented to get readers to somehow believe that it wasn’t written by an LLM. A deliberately entered error such as changing ‘are’ to ‘is.’ Nope. That doesn’t make us forget the constantly rolling ‘it isn’t just the’ / ‘it’s the’, ‘it doesn’t just’ / ‘it’, and ‘is less about’ / ‘is more about’ sentence combinations. I’ve just removed Comic Book Resources from my feed. <sigh>
