Wednesday, August 12th, 2026
Bill Bonner, from Poitou, France
AI is changing the world. But not necessarily you expect.
Like previous investment fevers — from the California Gold Rush to the Dot.com Bubble — it follows a well-trod path. The short version is this: speculators get excited...their early ‘investments’ raise prices...the companies spend money on chips and other ‘picks and shovels’...the spending is mistaken for end demand...and so much money stampedes in that earnings get trampled and can never justify the capital investment.
You can have all the intelligence you want — artificial, organic, instinctive...any kind you want — it’s not likely to change the pattern.
After the discovery of gold at Sutter’s timber mill in 1848, in rushed the miners, speculators, and suppliers. Claims were staked. Miners outfitted. With new money coming right up from California’s earth and streams, prices rose. In today’s dollars, a shovel cost more than $1,000. But then, when the easy surface deposits had been depleted, shovels were soon in over-supply.
And what’s this? Today’s nuggets — pulled from the dirt of our fake money credit market — are not as easy to get as they used to be. CNBC:
The 10-year Treasury note yield rose at the start of the week... The benchmark yield gained more than 4 basis points to 4.701%, while the 30-year Treasury bond traded up more than 3 basis points to 5.242%. The 2-year Treasury note yield rose more than 3 basis points as well to 4.241%.
Is this the beginning of the ‘bust’ phase for the AI Rush?
First, let’s look at the ‘theory’ underlying this massive bubble in AI-related investment.
The dot.com bubble was based on the idea that information could be transferred cheaply and quickly, electronically, thus boosting GDP. And with information so readily available, there was no excuse for people to make so many mistakes. Just search the internet for the right info. The ‘right’ answer was there somewhere. Eliminating bad investments would make the remaining uses of capital much more profitable. Like taking the dumb kids from the third grade...average test scores would go up.
It turned out that while the internet could deliver much more information and entertainment, it was promiscuous. Among the useful intel were millions of pages of drivel, click-bait, fake news, and idle entertainment. GDP growth, post 2000, went down, not up. Stupidity, bad taste and porno went up, not down. And did the internet avoid bad investments?
The worst single use of capital in the Internet Era was the attack on Iraq. Five trillion dollars down the drain. Anyone with an internet connection could have found that there were not likely to be any serious ‘Weapons of Mass Destruction’ in Iraq. Nor was Iraq a sponsor of terrorism. Nor did it pose a plausible threat to the US. But even access to 3,000 years of military history — disaster after disaster — did nothing to stop it.
So, now we put the question to AI...did the US have a good reason for its attack on Iraq? Claude:
No. Containment and inspections were working, and the war’s own advocates later conceded the central premises were wrong.
And now, we have a source not just of info, but of judgement — knowledge itself. LLMs (Large Language Models) are clever. They will do your thinking for you. No more ‘mistakes,’ right?
AI can now replace the bottom and middle rungs of ‘knowledge work.’ Those jobs would otherwise be taken by young, aspiring ‘white collar’ strivers, the very people whose earnings were supposed to keep federal debt from spinning out of control.
Also afoot are plans to bring in even more capital…and make AI even more accessible. Here is the summary from Grok:
On or around August 10, 2026, Nvidia announced memorandums of understanding with Apollo Global Management, BlackRock, Blackstone, Brookfield Asset Management, Goldman Sachs, KKR. The goal is to set up independent financing platforms that aim to mobilize more than $500 billion in third-party capital over time. This capital would help Nvidia’s customers (AI labs, enterprises, hyperscalers, and AI cloud providers) finance GPUs, systems, and data-center buildouts without relying solely on their own balance sheets
Whatever bone it is given to gnaw on, the value added is the ‘judgement.’ And there you see the fatal flaw.
AI has no divine source of truth. It merely rests, like an old boot on a trash heap, on words compiled over centuries, thoughts discarded by geniuses, lessons learned by con men and saints, visions that came at midnight and vanished by morning.
All human in origin...these judgments are subject to revision, interpretation, contradiction...
...and the boom/bust cycle.
More to come.
Regards,
Bill Bonner
Why is everything is so expensive? It’s been 64 months since the Federal Reserve last met it’s 2% inflation target. Average prices rose 3.4% in July, according to the Consumer Price Index (CPI) report released today by the US Bureau of Labor Statistics. The energy index fell 1.5% in July, with gasoline prices falling 2.9%. Over the last twelve months, the energy index is up 14.7%, with gasoline up 24.6% and fuel oil up 39.1%. According to US officials, the Strait of Hormuz remains ‘open for transit’.
Regarding Iraq: either there were no weapons of mass destruction, or Saddam sent them to Syria where they were used by Assad in subduing his people. Another possibility is that of all the Middle Eastern "countries" available, Saddam's Iraq was seemingly the easiest and most desirable mark. Like the stories of any modern American adventurism, for the benefit of the Deep State, no one of us knows the carefully guarded and manipulated truth. The same goes for political chicanery here at home: after 63 years, no one has paid for the skullduggery of the Kennedy Assassination, and despite many so-called releases of so-called classified information, we still do not have the full truth and nothing but the truth, save what Americans, on their own, have been able to figure out in the interim. This kind of corruption is a characteristic of what Mr. Bonner calls "late-stage decline of an empire." Best always. PM
LLM's are not intelligent. It's a machine that makes plenty of mistakes. Intelligence remains in the domain of humans. We just choose to not use it.