So, I am not sure about the tweet on the power consumption taking down the grid of a state. Tomβs Hardware analysis shows 3.5 million H100s would consume ~13 TWH power annually, which is about the same amount of power consumed annually by the city of Chicago. That is a ton of power. That is in the context of the US using 4,000 TWH compared to Chinaβs 6,800 TWH. This is great context - thanks for setting off this lightbulb!
They are gonna crash since LLM/neural nets' problems cannot be resolved and OpenAI and copycats have gotten AI stuck in a failed direction probably pushing back AGI decades, plus the general public absolutely hates AI and biotech and boos them at their own conferences these days.
Only thing propping up a lot of the new companies is endless Pentagon funding into AI, since it's an organisation of idiots who have created their own problem in search of a solution with more surveillance data than they can analyse.
US state probably doesn't have another 10 years before a civil war or revolutionary war or something else causing its collapse, so that's an unreliable source of funding in the near term esp. with global dedollorisation and loss of the dollar as the world reserve currency, largely propelled by their own policies.
Yes, it is a bubble. But we can capitalize on it in the moment. I'm not saying hold OpenAI stock (if they IPO) long-term. There is no way they will ever make their investment back.
LLMs are a dud like everyone who works with the software itself knows and AI is headed for a huge crash in the next year or so because of fundamental problems with transformers and next-token prediction and the underlying neural nets themselves, all of which has been known for a long time or even decades. Problems such as hallucinations and the inability to understand that A = B β΄ B = A cannot be resolved.
There is no market for LLMs, let alone a trillion dollar one, why do you think Sam Altman was going for a hail Mary pathetically begging for 7 trillion while he makes comments about how GPT5 won't be that good anyway? Nvidia has banked hard on AI, so it will probably crash around the same time the LLM bubble bursts. Meanwhile, AMD has only ever gotten better while Nvidia relies on exclusivity and anti-competitive practices just to get by in large segments of its market.
I very well know who JIm Simons is. If his method works then why would this fund be so tiny? It's pretty obvious to me what's going on at Renaissance Technologies and why the Medallion fund is capped at 10B but I don't want to spell it out because I don't want to get sued.
But thanks for commenting. I will look into your stuff too. Always ready to change my mind.
Good strategy... $NVDA is aiming for a 4 for 1 split and can get approval in June
Fingers crossed!
π―
So, I am not sure about the tweet on the power consumption taking down the grid of a state. Tomβs Hardware analysis shows 3.5 million H100s would consume ~13 TWH power annually, which is about the same amount of power consumed annually by the city of Chicago. That is a ton of power. That is in the context of the US using 4,000 TWH compared to Chinaβs 6,800 TWH. This is great context - thanks for setting off this lightbulb!
A datacenter takes much more power than the GPU alone. Servers, cooling, networking, and other things.
Great post!
Aswath Damodaran also has an interesting perspective on his analysis of Nvidia as well on his βmusing on marketsβ.
Unrelated to Nvidia, Iβve recently interviewed him if interested!
Very cool, will look into that.
The split has already drawn me to $NVDA for about 2 months ago. Outside crypto, Big tech and Big and emerging Pharma always rule the world!!!!
Look into AI and biotech. Some very interesting companies coming up.
They are gonna crash since LLM/neural nets' problems cannot be resolved and OpenAI and copycats have gotten AI stuck in a failed direction probably pushing back AGI decades, plus the general public absolutely hates AI and biotech and boos them at their own conferences these days.
Only thing propping up a lot of the new companies is endless Pentagon funding into AI, since it's an organisation of idiots who have created their own problem in search of a solution with more surveillance data than they can analyse.
US state probably doesn't have another 10 years before a civil war or revolutionary war or something else causing its collapse, so that's an unreliable source of funding in the near term esp. with global dedollorisation and loss of the dollar as the world reserve currency, largely propelled by their own policies.
Yes, it is a bubble. But we can capitalize on it in the moment. I'm not saying hold OpenAI stock (if they IPO) long-term. There is no way they will ever make their investment back.
Crypto means cryptography, not what you think it does.
I mean, it means both. I think we all know what they meant to say.
$smci will split before $nvda β¦
And whatβs your black swan? Have you ever managed other peopleβs money before?
I have managed other people's money for 20 years. I'm making a bet that $NVDA will skyrocket. Not financial advice.
$SMCI possibly you are right
What is your 2026 price target if you donβt mind me asking?
$SPY I said, βI will even go on record and say that when $NVDA hits over β1000.00β and $META β580.00β that would be the Marketβs Top.β
Did anyone listen?
βProbably Notβ ,β¦
LLMs are a dud like everyone who works with the software itself knows and AI is headed for a huge crash in the next year or so because of fundamental problems with transformers and next-token prediction and the underlying neural nets themselves, all of which has been known for a long time or even decades. Problems such as hallucinations and the inability to understand that A = B β΄ B = A cannot be resolved.
There is no market for LLMs, let alone a trillion dollar one, why do you think Sam Altman was going for a hail Mary pathetically begging for 7 trillion while he makes comments about how GPT5 won't be that good anyway? Nvidia has banked hard on AI, so it will probably crash around the same time the LLM bubble bursts. Meanwhile, AMD has only ever gotten better while Nvidia relies on exclusivity and anti-competitive practices just to get by in large segments of its market.
I think we all realize this is a bubble. Buy low, sell high. I don't think we are anywhere near the peak yet.
Great article! We analyzed the market movement of $NVDA. With/Without a stock split, it has still a way to go after the small correction wave. π
https://open.substack.com/pub/chartscope/p/unveiling-nvidias-technological-supremacy?r=1gn5e3&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true
Amazing stuff. Thanks!
I don't believe they work. I'm all in on fundamental analysis + news/psychology.
nobody ever proved pure technical analysis beats the market long term. But you're selling it, so hard to argue here.
I very well know who JIm Simons is. If his method works then why would this fund be so tiny? It's pretty obvious to me what's going on at Renaissance Technologies and why the Medallion fund is capped at 10B but I don't want to spell it out because I don't want to get sued.
But thanks for commenting. I will look into your stuff too. Always ready to change my mind.