#ai-bubble — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #ai-bubble, aggregated by home.social.
-
I think we're close to breaking them folks!
#ButlerianJihad #AI #AIBubble #TechBros #Technology #Google #AIArt #AIMusic
-
-
You may see the title of this discussion and think "I've heard this all before". You might be mostly right about that. But I still strongly recommend listening to this, because the interview is long enough (unlike the tiny sound bites we're inundated with) that there's time to put a lot of things (the AI bubble and the motivations of those buying into it, Silicon-valley enshittification, decades of antitrust failures) into place in a coherent perspective.
-
Software was never insurable...
...it's idiots all the way down.Edit: Ah yes, the Shroedingers #Ai #datacenter
Simultaneously rushing headlong to #aibubble AND making 1 billion in profits."An average 100-megawatt data center, for example, could generate more than $1 billion in annual profits,"
It's idiots all the way down
-
I am concerned that we are maximizing energy consumption while systematically wiping out every method to safely provide for it. #EDF #energy #AIBubble https://www.edf.org/media/trump-administration-cancels-major-electric-transmission-corridors-intended-curb-costs
-
The AI boom: rational enthusiasm or the next dot-com bubble?
-
If anyone is using Anthropic AI models, I would suggest not to.
There is speculation that enterprise focused Anthropic is rushing to do an IPO for trillions in valuation to overshadow spacex (like another marketing stunt), which would give them overwhelming resources to continue blow the AI bubble (with these new influx of vast sums of money) and concentrate power even more.
IF you use AI - don't pay them money (*with exceptions, because they loose a lot of money on power users), and don't add to their revenue for their investors to show how great they are doing.
They are not doing that great, as small Chinese open models have significantly improved in the last few months and can be used instead to achieve similar results for a lot less money.
Such as "Deepseek v4 Flash 0731" 250B size model, which costs 62X 😯 less than Anthropic best (available) model, and which also has higher intelligence than Anthropic's best available model from 6 months ago.
Or recently released "Qwen 3.8max" 27B small model variant, which can be run on the desktop locally, and is also more 'intelligent' than Anthropic's best model was 6 months ago.
Models don't have to be big, they can be smaller. One does not need to transfer Anthropic money for their best models, because these fresh small models can do the the job*.
*With exceptions - small models have a lot of distilled intelligence for their size, but they don't have a lot of knowledge. When some task require some deep-niche-specialized knowledge (like translating between small languages), then one can try use large variants like "Qwen 3.8max" 2.4 trillion parameters sized open model. It cost a lot less than Anthropic's as well.
You can find prices on OpenRouter https://openrouter.ai/ to choose from countless providers of open models (but OpenRouter is also becoming big tech right now, it was recently was bought by Stripe for 7 billion), so there are also French alternative AI router https://www.edenai.co/, or UK alternative AI router https://www.requesty.ai/, or privacy-crypto-chat-UI AI router https://nano-gpt.com/ to connect to open model providers). And for agentic coding software (so called "AI agent harness") use https://kilo.ai/ open source software (or choose from many other options), not "Claude Code".
And for everyday (non-coding) tasks, one can choose privacy oriented OpenAI alternative @protonprivacy Lumo with regular chat interface. They are using very capable open AI models as well.
I attached visualization how models evolved recently, and compared where Proton Lumo 2.0 currently stand.
And if you don't use AI, so continue to not use it. But if you do, these better alternatives are available. "Better" does not mean "right" - open weights AI can also become very problematic. It is at least better than to support AI companies of very concentrated power.
If movement to open models spread quicker than Anthropic IPO plans, hopefully Anthropic will go bankrupt for good. #aibubble #LumoAI
-
My thoughts on #LLM :
1. I can see useful, ethical use cases for small-scale, specially-trained #LLM systems.
2. However, none of the big #AI companies - #Antropic, #OpenAI, #SpaceX , #Amazon, #Microsoft and so forth - are acting in an even remotely ethical manner when it comes to training their LLM systems.
3. Furthermore, I _really_ don't see the trillion dollar business case for "Let's scrape every website on the Internet a thousand times per month in the hopes of finding new data!"
4. That scraping is also pissing me off personally since I had to upgrade my hosting plan _twice_ and _still_ get the occasional outage from all that scraper bot traffic.
5. Therefore, I will not use any of the AI models by the current crop of corporate overlords.
6. Instead, I will wait until the #AIBubble inevitably pops and the world economy comes crashing down.
7. And after this happens, then _maybe_ there will be something worth using after we have sifted through the ashes. -
Sometimes I think causing the #RAMpocalypse and taking compute out of everyday hands was the whole point of the #AIbubble. Certainly makes more sense than everything else AI.
-
“None of this has much to do with my feelings about #AI, and far more to do with basic mathematics. #OpenAI has no economies of scale, it’s horribly-unprofitable, and does not have a stable business.”
-
En Europe aussi, « l'offre de titres de dettes souveraines s’ajoutent celles des entreprises et notamment des géants de l’intelligence artificielle (IA). "Les hyperscalers lèvent des montants faramineux pour financer leur développement. Les volumes recherchés se rapprochent de ceux des Etats, mais les taux sont plus élevés et donc plus attractifs aux yeux des investisseurs". » https://www.lemonde.fr/economie/article/2026/08/20/comment-les-marches-financiers-font-bondir-les-taux-souverains-europeens_6750402_3234.html #AIbubble
-
That being said, I think it's a good idea to digitised everything we can (without damaging the physical media, of course, even if it's more expensive) and make it publicly available. Every piece of knowledge and information, every work of art and culture, everything should be as visible and accessible as possible. Filesharing shouldn't be prosecuted but encouraged.
While I think that inflating LLMs to absurd sizes is absolutely mental, I think that machine learning is one of the greatest inventions of the computer age, even if that age probably won't last for very much longer due to the collapse of our high-tech civilisation. The problem isn't AI, the problem is capitalism. AI research and development should be a cooperative open source effort with openly accessible training data, and the resulting models should be in the public domain, owned by nobody in particular and free for everyone to use as they see fit.
If you only think "chatbots" when you hear "AI", that's because of marketing. Of course AI is just a marketing term (always has been), it's an umbrella term used for a whole lot of different branches of computer science which have only one thing in common, which is that they aim at solving complex tasks with computers for which we humans use our intelligence. One of those branches is machine learning, of which LLMs are just one application. While LLMs are quite fascinating from a research POV, they are highly experimental and not even worth a fraction of the money that has been wasted on all those big AI companies like OpenAI or Anthropic. We should study them and see what we can learn from them, see how we can use what we learn from LLMs for future AI R&D, and look into how small and compact we can make LLMs while still retaining functionality for the few use cases where LLMs are actually useful.
Whether we like it or not, AI won't be going away as long as we still have computers. The big AI companies probably won't last because they haven't got any viable business model, their LLMs are far too huge and inefficient, the costs of training and running those models are too high, and there aren't any users willing or able to pay for that. However, there are many small-ish LLMs which you can download and run on your own machine locally, and even after the #AIBubble bursts, they will remain, many nerdy people with big fat PCs will download and use them. Also, image generators, video generators, natural sounding speech synthesis models, etc., are also quite often small enough to run locally on a decent PC, and many, many people are using them. The open weights models can be fine-tuned and modified with LoRAs, which are like patches for artificial neural networks, and they are often used by AI hobbyists to decensor censored models, so AI generated pr0n isn't going away, either.
Most AI generated pornography is quite silly though. AI threads on 4chan /gif/ often have scenes from TV shows that suddenly turn into porn, like Counsellor Troi suddenly undressing on the bridge of the Enterprise, and then Riker gets his dick out, and Troi sucks it. Things like that exist in our world now, Rule 34 has completely exploded due to generative AI. And of course all kinds of people use it for propaganda and misinformation, although unfortunately, 99% of that comes from the extreme right, whether it's fascists or religious extremists.
When I was a teenager in the year 1990 and had just acquired my first SuperVGA card that could do colour in 24 bits, I absolutely expected to see the day when everyone could just create images and videos that looked absolutely real, destroying the trust in visual media where it still existed. CGI in movies and on TV was just taking off, but I could see where it was going. For the next two decades, creating something that looked natural and convincing still took a lot of knowledge, skill, and effort, even with the newest digital tools. In the 2010s, generative AI began to change that. Now we live in that world I saw coming as a teenager, where we can never again trust anything we see on a screen. There isn't much we can do about that. Of course we can force AI companies to watermark their digital output, but we cannot force people running AI on their own machines to do that, and secret propaganda outlets, whether run by states or by the private sector, obviously aren't going to do that. The only thing we can do is teach people not to trust anything they see on any screen. Or in print, for those of us who still read newspapers and magazines.
Using the rubbish idea of "intellectual property" to fight against generative AI is completely bonkers. Intellectual property wasn't invented to protect creative workers and artists, it was invented to protect capital investments. Copyright was invented to protect English printers against their Scottish competitors in the 1800s.
-
Pine64 takes a knee while waiting for the clock to run out: noteworthy.
I was one of the original purchasers of the pine64s when they were just starting. Made a beowulf cluster with 10 of them.
So a bit of sad.
-
RE: https://graphics.social/@metin/117123760167000114
Seems like any nation going all in on AI at every level is really shooting themselves in the foot.
That graph really hammers home the AI Brain Rot
-
Le toujours formidable Cory Doctorow aka @pluralistic à propos de la bulle de l'ia et du risque fasciste de la crise économique qui vient
#ia #ai
#nightmareOnLLMstreet
#Aibubble
#LateStageCapitalism
#Fascism
🌍$💀 -
ECB economists warn households to brace for stock market ‘correction’
https://www.independent.ie/business/ecb-economists-warn-households-to-brace-for-stock-market-correction/a/160151547.html
Booming AI shares have pushed markets to levels where a sharp fall is now likely, according to a stark warning issued by European Central Bank (ECB) researchers. -
Jim Cramer: market negativity is creating buying opportunities!
John Stewart: (sharing video of a certain March 12, 2009 interview with his writers...)
#ai #aibubble #privateequity #tariffs #iranwar #tankbottom #Bondyields #inflation #stagflation
#frogswarms #lice #flies #pestilence #hail #locust etc
https://www.cnbc.com/2026/08/18/jim-cramer-market-too-negative-buying-opportunities.html
-
“OpenAI on Tuesday said it had slowed down the pace of its #AIDevelopment while it overhauled its research and training systems.
The company’s researchers were caught unaware last month when an AI agent under testing hacked another AI firm, #HuggingFace.
The AI research lab behind #ChatGPT said its new measures included pausing its model testing for two weeks and investing more in adding other #AISystems to monitor the activities of AI agents in testing. Some of the company’s largest planned training runs remain on hold, the company said.”
One way to say, “finances are tight.”
#OpenAI / #AIBubble <https://theguardian.com/technology/2026/aug/18/open-ai-pause-hack>
