Comment by jrflo
10 hours ago
Thanks for putting this together, his writing really aggravates me and it was nice to see all of his predictions put together in one spot.
It's interesting how he has pivoted from denying that LLMs are useful to accusing frontier labs of Enron-scale fraud.
I suppose it's the natural pivot you have to make when your whole business is an anti-AI newsletter that costs $7 per month.
The fraud is there. Whether it surpasses Enron or not will be revealed in the future.
To commit fraud you have to lie, deceive, or conceal. All of this bonkers financing is happening in broad daylight, announced to the markets, and met with rave reviews. I tend to agree with Zitron that the whole thing is a castle made of sand - not because the tech is bad or useless but because the financing, in its scope and structure, is so outlandish - but you can build a castle made of sand as long as you don’t tell anyone it’s made of steel.
So far NVIDIA has been very careful, it’s not actual fraud, at least based on public information. It’s a very unstable and extremely risky bet, that is very likely to blow in the face face of AI companies, but it’s not fraud
I think a lot of people (myself included) would love for it to be fraud. Because even if it's not fraud what is being done between NVidia, OAI, Anthropic, Microsoft, Meta, Musk and Google is funny accounting at the very least. It's hard to take any part of what's going on seriously in the US at this point and I feel like the bigger downside to all of this is that if/when they get away with this, and there are no repurcussions for any of the big names involved, then it erodes consumer trust that much further, pushing us deeper into a state of "doesn't matter anyway, why try".
2 replies →
Patrick Boyle had a vid specifically on whether there was Enron style fraud. The conclusion was no, though he thinks things may be a bit bubbly https://youtu.be/NufJ7g63KSY
It's worth considering that it might not be 'Enron-style' fraud, because that's the best comparison have at the moment.
What event did people compare Enron to, before Enron became 'Enron'?
It's probable we're witnessing an entirely new fraud, we just don't have all of the details yet.
1 reply →
There is no evidence of fraud that I have seen beyond complete speculation from the likes of Zitron. He basically just says "Look how big the numbers are! It's so big it must be fraud!"
Zitron himself doesn’t call it fraud and doesn’t compare it to Enron (he said no, with his reasoning when asked in interviews, multiple times)
3 replies →
He LITERALLY said it is not like Enron. And repeated it.
I was thinking of this article where he compares OpenAI to Enron twice: https://www.wheresyoured.at/what-happens-if-openai-dies/
You're right that he doesn't explicitly say enron-scale fraud, that's just what I came away with from reading the article a while back.
I haven't read every word but I don't recall him saying AI is bad or useless ever either. His refrain is that they cannot keep up with their endless spending on training and they're not reaching new markets.
I look at the rapid rise and decline of OpenClaw and get the impression AI has not really found any foothold among average people.
From the OP:
August 2024: "generative AI is a dead-end technology that has peaked”
July 2024: "Generative AI models aren’t getting more energy-efficient, nor are they getting more “powerful” in a way that would increase their functionality"
Nov 2025: "the fact we're running out of high quality training data and we're hitting the walls of scaling laws, in the training paradigm, these models aren't getting better. What we're seeing today is pretty much what they're always gonna be like"
There are other examples too from his prose of talking about how they are barely useful but I don't want to dig it up
> I don't recall him saying AI is bad or useless ever either.
He does say that pretty often in interviews. That doesn’t change his thesis but that‘s one reasonable reason people dismiss him, he has often said that AI is useless when considering the externalities. And he will sometimes take a shortcut and just say „it’s useless, doesn’t do anything well“, which is of course way too simplified.