Lost thoughts

The is where I am, lost in thoughts. Drawn between my personal conviction that the AI bubble is real and the set fake thoughts on LinkedIn and Youtube making ‘their’ case on the AI bubble. One is set on thoughts of doubts considering the technology we are currently at, the other thoughts are all fake perceptions by influencers trying to gain a following. So how can any one get any thought straight? Yet in all these there are several people in doubt on their own set (justified) fringes. One of them is ABC who gives us ‘US risks AI debt bubble as China faces its ‘arithmetic problem’, leading analysts warn’ (at https://www.abc.net.au/news/2025-11-11/marc-sumerlin-federal-reserve-michael-pettis-china/105992570) So in the first setting, what is the US doing with the AI debt? Didn’t they learn their lesson in 2008? In the first setting we get “Mr Sumerlin says he is increasingly worried about a slowing economy and a debt bubble in the artificial intelligence sector.” That is fair (to a certain degree) a US Federal Reserve chair contender has the economic settings, but as I look back to 2008, that game put hundreds of thousands on the brink of desperation and now it isn’t a boom of CDO’s and stocks. Now it is a dozen firms who will demand an umbrella from that same Federal Reserve to stay in business. And Mr. Sumerlin gives us “He is increasingly concerned about a slowdown in the US economy, which is why he thinks the Fed needs to cut interest rates again in December and perhaps a couple more times next year.” I cannot comment on that, but it sounds fair (I lack economic degrees) and outside of this AI bubble setting we are given “US President Donald Trump has recently posted on his social media account about giving all Americans not on high incomes, a $US2,000 tariff “dividend” — an idea which Mr Sumerlin, a one-time economic adviser to former US president George W Bush, said could stoke inflation.” I get it, but it sounds unfair, the idea that an AI bubble is forming is real, the setting that people get a dividend that could stoke inflation might be real (they didn’t get the money yet) but they are unrelated inflation settings and they could give a much larger rise to the dangers of the AI bubble but that doesn’t make it so. The bubble is already real because technology is warped and the class cases we will see coming in 2026 is base on ‘allegedly fraudulent’ sales towards the AI setting and if you wonder what happens, is that these firms buying into that AI solution will cry havoc (no return on AI investment) when that happens and it will happen, of that I have very little doubt. 

So then we get to the second setting and that is the clam that ‘China has an arithmetic problem’, I am at a loss as to what they mean and the ABC explanation is “But if you have a GDP growth target, and you can’t get consumption to grow more quickly, you can’t allow investment to grow more slowly because together they add up to growth. They’re over-invested almost across the board, so policy consists of trying to find out which sectors are least likely to be harmed by additional over-investment.”

Professor Pettis said that, to curry favour with the central government, local governments had skewed over-investment into areas such as solar panels, batteries, electric vehicles and other industries deemed a priority by Beijing.” This kinda makes sense to me, but as I see it, that is an economic setting, not an AI setting. What I think is happening that both USA and China have their own bubble settings and these bubbles will collide in the most unfortunate ways possible. 

But there is also a hindsight. As I see it Huawei is chasing their own AI dream in a novel way that relies on a mere fraction of what the west needs and as I see it, they will be coming up short soon, a setting that Huawei is not facing at present and as I see it, they will be rolling out their centers in multiple ways when the western settings will be running out of juice (as the expression goes). 

Is this going to happen? I think so, but it depends on a number of settings that have not played out yet, so the fear is partially too soon and based on too little information. But on the side I have been powering my brain to another setting. As time goes I have ben thinking through the third Dr. Strange movie and here I had the novel idea which could give us a nice setting where the strain is between too rigid and too flexible and it is a (sort of) stage between Dr. Strange (Benedict Cumberbatch) and Baron Mordo (Chiwetel Ejiofor) the idea was to set the given stage of being too rigid (Mordo) against overly flexible (Strange) and in-between are the settings of Mordo’s African village and as Mordo is protecting them we see the optional settings that Kraven (Aaron Taylor-Johnson) get involved and that gets Dr. Strange in the mix. The nice setting is that neither is evil, they tend to fight evil and it is the label that gets seen. Anyway that was a setting I went through this morning. 

You might wonder why I mentioned this. You see, Bubbles are just as much labels as anything and it becomes a bubble when asset prices surge rapidly, far exceeding their intrinsic value, often fueled by speculation and investor orgasms. This is followed by a sharp and sudden market crash, or “burst,” when prices collapse, leading to significant rather weighty losses for investors. And they will then cry like little girls over the losses in their wallets. But that too is a label. Just like an IT bubble, the players tend to be rigid and whole focussed on their profits and they tend to go with the ‘roll with it’ philosophy and that is where the AI is at present, they don’t care that the technology isn’t ready yet and they do not care about DML and LLM and they want to program around the AI negativity, but that negativity could be averted in larger streams when proper DML information if given to the customers and they dug their own graves here as the customer demands AI, they might not know what it is (but they want it) and they learned in Comic Books what AI was, and they embrace that. Not the reality given by Alan Turing, but what Marvel fed them through Brainiac. And there is a overlap of what is perceived and what is real and that is what will fuel the AI bubble towards implosion (a massive one) and I personally reckon that 2026 will fuel it through the class actions and the beginning is already here. As the Conversation hands us “Anthropic, an AI startup founded in 2021, has reached a groundbreaking US$1.5 billion settlement (AU$2.28 billion) in a class-action copyright lawsuit. The case was initiated in 2024 by novelist Andrea Bartz and non-fiction writers Charles Graeber and Kirk Wallace Johnson.” Which we get from ‘An AI startup has agreed to a $2.2 billion copyright settlement. But will Australian writers benefit?’ (At https://theconversation.com/an-ai-startup-has-agreed-to-a-2-2-billion-copyright-settlement-but-will-australian-writers-benefit-264771) less then 6 weeks ago. And the entire AI setting has a few more class actions coming their way. So before you judge me on being crazy (which might be fair too) the news is already out there, the question is what lobbyists are quieting down the noise because that is noise according to their elected voters. You might wonder how one affect the other. Well, that is a fair question, but it hold water, as these so called AI (I call them Near Intelligent Parses, or NIP) require training materials and when the materials are thrown out of the stage, there is no learning and no half baked AI will holds its own water and that is what is coming. 

A simple setting that could be seen by anyone who saw the technology to the degree it had to. Have a great day this mid week day.