The Algorithm That Detected a $610 Billion Fraud: How Machine Intelligence Exposed the AI Industry’s Circular Financing Scheme
The Algorithm That Detected a $610 Billion Fraud: How Machine Intelligence Exposed the AI Industry’s Circular Financing Scheme
Kind of gross how this article seems to be trying at every turn to say, "no ai is actually good! It helped us catch the bad businessmen that happen to be in the AI industry!" By focusing on a tiny trading period on November 20th.
Hank Green isn't a finance bro or an AI guy or even really a tech guy. He's just a guy reacting to things that are trending, and I remember I had seen the main graphic he was talking about floating around the internet for a while before I watched the video. People have been calling AI a "bubble" for much longer.
I am old enough to remember the report that 95% of generative AI companies failed to see returns from using it. That was back in August.
I don't like giving credit to "trading algorithms" for things that humans figured out a long time ago.
There was a meme going around a month or so ago of 5 AI companies jerking off, representing the circular financing that this article talks about.
From the moment these "investments" were announced, people were already raising flags about circular investments.
I feel like you're kinda underselling Hank Green a bit. I mean he is the CEO of like a million companies. But it's fair to say he doesn't have a finance background at the very least.
Oh yeah I have as much respect for him as I can have for any other celebrity I've never met or interacted with at all. I just wanted to get ahead of anyone responding to me pointing out that he's not particularly qualified on this subject.
The reason I referenced him at all was not because of his qualifications, but as a way of establishing how popular these conversations were on the internet at the time. In that sense, the fact that he's not an AI or finance expert and doesn't specialize in such content speaks to how widespread the topic was at the time.
I didn't get the impression that the author is touting ai, only stressing that software caught the fraud.
Don’t say AI found it, just an algorithm caught it.
Could have been as simple as a spreadsheet highlighting cells with a ratio that was below a threshold.
Article didn’t say what software was used.
Edit: I meant “didn’t say”. Autocorrect has been really going after me lately.
But the AI or algorithm didn't catch it - humans did just fine with that part.
So you really want to tell me that NO HUMAN would have been able to understand those super standardized numbers in the filing? Sure man.
It took 18h for the AI to detect this. 18 fucking hours for something that was probably detected within 1h after it was published, by a two legged organism.
The article said it was an algorithm that detected it, not AI.
Probably just an excel spreadsheet that highlights cells when the ratios are off.
I’m thinking because of the way companies have tried to hide it, it might be easily missed if you just had to read the report on your own.
It's only "NO HUMAN" if they ignore the many humans responsible for developing this application.
But that's totally normal in the griftosphere.
It’s only “NO HUMAN” if they ignore the many humans responsible for developing this application
Or the fact that humans already posted a video about this circular financing a month earlier
And don't forget that the two legged organic machine would have used way less energy to do so
I am not so sure since I saw Super Size Me.
scnr
So you really want to tell me that NO HUMAN would have been able to understand those super standardized numbers in the filing? Sure man.
Gamers Nexus managed to figure it out a month earlier using good old-fashioned HI.
I am so glad that we have people like Stephen Burke on this planet. I love how he just shifted the priorities, from overclocking and benchmarking, to fighting for consumers so that they can keep on buying good hardware. He deserves much love.
Don't they call Burry the GOAT?
On November 20, 2025, trading algorithms identified what may become the largest accounting fraud in technology history
Did these algorithms just look up the Gamer's Nexus video about the circular financing that was published a month earlier?
Financial analysts were sounding the alarm in October. On October 7, Bloomberg ran an influential article about the circular deals:
That built on earlier reporting where they described the deals as circular, as the deals were being announced. Each of these reports notes the financial analysts at different investment firms sounding the alarm.
From there, a robust discussion happened all over the financial press about whether these circular deals were truly unstable. By the time Gamers Nexus ran that video the financial press was already kinda getting sick of the story.
Whatever the hell these trading algorithms were doing on November 20, they definitely weren't ahead of the curve on investor knowledge and belief.
Whatever the hell these trading algorithms were doing on November 20, they definitely weren’t ahead of the curve on investor knowledge and belief.
Yup. But you'd need to keep your head out of the AI bubble's arse to notice that. :D
Looks more and more like a vulgar Ponzi scheme. Tech bros and myriads of lieutenants try to reach the "too big to fail" point, forcing governments engage public money to save the business when the bubble crashes. Brilliant.
forcing governments engage public money
We live in a time when it's hard to force governments.
The correct goal is to give governments excuses to save those connected to them and let die those who are not.
USSR's breakup and the dotcom bubble started a new era. Everyone saw that this works and there's no higher wisdom or hidden fallback mechanism to prevent this.
But the incentive to "save businesses" is, well, why I'm a libertarian most of the time. If governments had no money to be directed by some central strategy, and instead only the means to minimally function as publicly decided - fire services, infrastructure, electoral procedures, IDs and money, - then there would be no option to involve them in such a way.
And it makes sense that libertarians are usually very vulnerable to the AI hype, as to the cryptocurrencies hype before it, an irony typical for history. That means that they'll be those hit the most.
It's an intended minefield. There's a road called "techno-optimism and individualism", and most of the mines are being laid on it. Similar to the KGB "rotten herring" thing and such. To discredit an idea. It's more believable when the splattered meat around those already exploded is real.
Again, look at USSR, its ideology had plenty of flaws, some with pretty infernal consequences, but it was the main one putting future united humanity, progress, science, equality, humanism and secularism into the center of its cosmology. It offered pretty dubious tools, but that's irrelevant. When it failed, all those things listed also got a hit. It's not a coincidence that "polemical" (with both dystopia and utopia and questions about human nature put together in the same space) science fiction in the 90s transitioned into clearly dystopian "putrid" "dream denied" cyberpunk.
but it was the main one putting future united humanity, progress, science, equality, humanism and secularism into the center of its cosmology
Just llike with many people, what they say and what they do can be two very different things.
In the USSR's case, they talked socialism while being totalitarian state capitalists. This was even noted at the time of the Russian revolution by people such as Rosa Luxemburg. And in terms of its foreign policy, they talked internationalism and cooperation while carrying on the legacy of Russian imperialism.
When it failed, all those things listed also got a hit.
All those things also took a hit at the end of WW1, in the Great Depression, and during WW2. Nobody who was paying attention believed the Russian bullshit, before or after the collapse of the Communist Party's rule. The double-dealing in the Spanish Civil War, the Molotov-Ribbentrop pact, the crushing of the Prague Spring, and many other events put paid to that naivety.
If governments had no money to be directed by some central strategy,
You mean the central strategy of creating the money, imposing capitalism, and violently enforcing extreme privilege/deprivation?
it makes sense that libertarians are usually very vulnerable to the AI hype
Because they're already disconnected from reality and worshiping an obvious pseudo-science?
they are aiming for that BAILOUT money, thats why they have been greasing the gop and trump for years.
I'm pretty sure I watched some random youtuber's video explaining how circular this shit is nearly a month ago... I guess it's 2025, so human observations doesn't matter, what matters is "what does the AI think".
You are correct. Gamers Nexus did a video on this a month earlier.
Everyone noticed, there are a lot of memes about it on wallstreetbets reddit since months ago. This is old news.
why call Ed Zitron an algorithm tho
I love Ed. He is the personification of my emotions.
Yeah I'm glad this news is like everywhere and a major scandal and there will be a fallout
There will be fallout, for sure, as soon as they finish calculating what percentage of profit should be fined away.
ELI5?
If I gave you $5 and then you gave it to someone else and then they gave it back to me we've done nothing but can call it $15 in business transactions.
Nvidia invests in company.... Company buys Nvidia items.... Nvidia stock goes up.... Nvidia has new pretend money to invest into another company....
That's why transactions are defined as the exchange of something of value. In most legal systems, if you're caught swapping phony transactions like that, you can be prosecuted for fraud.
I think I understand. Can... Can I have the $5 now please?
300% profit! Where's my bonus!
I haven't read the article, but I have read previous accusations of the same thing, so I assume it's the same.
Basically, the new AI companies are all losing money, but they are all investing big money in each other which makes it look like the industry is doing well.
I think this snippet get the gist across:
The money flows in loops: Nvidia invests in AI startups, startups commit to cloud spending, cloud providers purchase Nvidia hardware, Nvidia recognizes revenue, but the cash never completes the circuit because the underlying economic activity—AI applications generating profit—remains insufficient.
"Every public company now faces machine-speed scrutiny of accounting practices. Anomalies that might have persisted for quarters until human analysts identified patterns now trigger immediate algorithmic responses."
I like how there are all these terms with increasingly loose definitions, to which we attach different levels of evilness:
Machine intelligence and aritificial intelligence are the same thing. AI just changed its meaning too much i guess. What was AI few years ago, is now machine intelligence.
(Somebody correct me if i'm wrong, i know you want to...)
The original comment or first line was
"I like how there are all these terms with increasingly loose definitions, to which we attach different levels of evilness"
On a technical level, I agree machine intelligence and artificial intelligence are closely related, if not the same definition. However, they are not synonymous for the average person, going back to the "loose definitions" bit.
Good reading. Besides Nvidia and the AI money bubble, the link to Bitcoin is interesting.I‘ve read so many other articles about financial acrobatics with Bitcoin and how it collapses now, I‘m waiting to see it falling down to 50k.
I‘m waiting to see it falling down to 50k.
Yeah, if you look at long-term averages, this is a realistic buy price.
"AI" <---> "AI"
Meaningless grift fixed by meaningless grift? Sounds like peak grift.
Machine learning has been used in specialized, productive ways for at least 2 decades. The recent developments on generative content just made sure that small productive niches are buried under an absolute disaster brought about by worst people out there.
I don't know how to tell you this, but not everything a computer does is AI.
But it is though.... AI was started long before computers were made.
AI at its core means "intelligent behaviour exhibited by non-natural/Artificial means". This means that if your OS "intelligently" switches tasks on physical threads on cpu that's AI. If you go to supermarket and the door opens when you go near it, that's AI as well (actually this example was in our intro to AI textbook).
So computers are a special case of AI, so are calculators.
AI != Al
One has a lower case L. But most people don’t check for fontjacking and I just sold a 3 billion data center for Al 👍
So AI exposed AI scam?
So uh AI works then and will fix all our problems that it caused in the first place 👍
Roflmao
Is this what he meant by we should be using AI for everything?
Absolute idiots
The rest of us knew it was a scam all along. and we didn’t need AI to figure that out.
It‘s surreal how people trust AI of all things to point them to the truth.
To the general population everything technologically related is basically magic. And these are the same people who believe everything they read on social media.
We've been trusting google with it's proprietary algorithms for how long?
We also trust politicians, business leaders with PR teams crafting their every speech and press release...
We also all trust Google, Apple, Microsoft and many other companies with all of our data and metadata. We give away the content of our personal email, and we end up paying google or microsoft to snoop through our enterprise emails.
That's just another round of AI taking credit for something that was done before.
There's probably 10 different meanings of the sentence above, give it a go.
It's cause we powered our human intelligence with Twix and cocaine!
While true, what actually happened was interesting. Algorithm's are not inherently bad, afterall.
At some point, someone wrote a rule to verify DSO against other tech companies, and trigger an automated short position to correct the market changed from an earnings call.
To me, that's pretty cool. This wasn't magic LLM AI, this was a smart engineer that programmed a system to discover problems as they arose.