The LLM Bubble is About to Implode, Gimmicks and Financial Shell Games Cannot Prevent That, Only Delay It
Earlier today, under "copyrights" and "proprietary" in our Daily Links [1, 2], we added some news links [2-4] (see references at the bottom) that neatly related to many gloomy predictions associated with financial problems at Microsoft OpenAI (there's lots more of the same; but a few may suffice for now). It's running out of money, even the money it has loaned. It's not just Microsoft OpenAI though [1], it's all this nonsense that adds "generative" or "advanced" or other buzzwords to "AI", which has itself become an adjective, not an acronym, added to describe things that aren't new and aren't Machine Learning. There is almost nothing novel here, only brute force 'on tap'. Someone pays the bills for useless "games" with no solid use cases. Not many people would actually pay for this (some who did demand a refund or sue for a refund).
To inflate the bubble MElon is now doing the classic trick of buying from oneself for a fictional value [5]. That's the best they can do at this point. Microsoft came under investigation for it.
Regarding this fake news ("X merges with xAI"; whatever that even means), one reader wrote to ask: "What do you think of Xai's purchase of X (former Twitter)? Both firms are controlled by Elon Musk. So it may be something like moving coins from one hand to another."
Very much like debt-loading. Microsoft is trying to hide its financial problems by offloading them to other shells, which can be dissolved rather than liquidated, i.e. burn debt. We know at whose expense. An associate has twice said that "Microsoft OpenAI and Scam Altman continue to steal Ghibli's effort, and use that as a distraction from Microsoft OpenAI's economic unviability" (and business that has failed to retain users; the engagement has waned, contrary to phony numbers floated by the broke company).
It would be nice to see this LLM bubble coming to an end. Scam Altman trying to enter the "slop image" market is embracing a bubble; many of those companies cannot find a business model and already shut down their entire operations. They attracted many lawsuits already.
If "slop images" and "slop text" are going to become harder to manufacture (via a form of plagiarism services), then maybe there's a chance for the Web to improve and even recover somewhat. It would relieve many sites; LLM scrapers and slop image scrapers stage DDoS attacks, whether willfully or not. They're so greedy and selfish; they lack intelligence.
Many news sites threw in the towel not just due to plagiarism bots but also the DDoS-like behaviour, which increases the cost of operating vital sites.
"In 2003," the associate recalls, "it wasn't even remotely feasible to keep up with just the headlines let alone articles. Thus sites like the late, great Slashdot were so important then. Before they were sold, they captured the zeitgeist: FOSS."
"However, that ground some gears high up and each round of new owners stifled that further and increased the anti-FOSS propaganda an notch."
"Now you have Conde Nast's Reddit which all to many youth mistake for a discussion site."
Reddit is selling data to LLM sloppers.
Worse yet, an hour ago I found out that some sport articles from official sites are slop and may contain image slop, too. Imagine how bad it would be if CBC, BBC, ABC etc. let us all ingest chatbots. People would simply not bother anymore.
Let's hope this bubble implodes as soon as possible. Baidu’s CEO "believes 99% of AI companies won’t survive after the bubble bursts," this report said some months ago.
Are we there yet? No. But we're getting there. █
_______
-
Daniel Miller ☛ Comments on How Software Engineers Actually Use AI
I’m much more concerned about the economic impacts of the current AI bubble. I lived through the first dot-com bubble, and I was too young and too early in my career to appreciate its broader impacts (even though it led to the loss of my first job in the software business). The amount of capital currently being poured into what is essentially an unproven technology — at least from the perspective of proving some massive value that would justify the investment – is frankly terrifying. Cory Doctorow has written the most cogent analysis I’ve read thus far in What Kind of Bubble is AI?: [...]
-
Variety ☛ Studio Ghibli Distributor Praises 'Princess Mononoke' Release Over OpenAI
Hollywood actors and other creatives have voiced concerns about efforts by OpenAI and other artificial-intelligence companies to “weaken or eliminate” protections on copyrighted works for training AI systems. In comments filed with the Trump administration‘s Office of Science and Technology Policy earlier this month, more than 400 filmmakers, actors, musicians and others objected to what they said was lobbying by OpenAI and Google “for a special government exemption so they can freely exploit America’s creative and knowledge industries.”
-
The Independent UK ☛ Studio Ghibli distributor pushes back at AI attempts to ‘replicate humanity’ as classic film returns to cinemas
The trend has delighted fans but also highlighted ethical concerns about artificial intelligence tools trained on copyrighted creative works and what that means for the future livelihoods of human artists.
-
Hindustan Times ☛ Who is Hayao Miyazaki? Ghibli co-founder's rant against ‘utterly disgusted’ AI animation goes viral
Thousands of people utilized OpenAI's picture generator in the GPT‑4o model to generate Studio Ghibli-inspired art on social media. However, critics and admirers of the renowned Japanese animator Hayao Miyazaki have voiced their displeasure of the trend and copyright issues.
-
Nick Heer ☛ Elon Musk Gives Himself a Handshake
This feels like it has to be part of some kind of financial crime, right? Like, I am sure it is not; I am sure this is just a normal thing businesses do that only feels criminal, like how they move money around the world to avoid taxes.