Click here: to donate by Credit Card
Or here: to donate by PayPal
Or by mail to: Free Republic, LLC - PO Box 9771 - Fresno, CA 93794
Thank you very much and God bless you.
Posted on 02/13/2026 12:25:39 PM PST by SmokingJoe
It's uploaded by the channel Dr. Josh C. Simmons on February 13, 2026 (today, based on current timing), and has quickly gathered around 5,285 views and 323 likes so far. Key Points from the Video
The core argument is that major tech companies (Microsoft, Alphabet/Google, Amazon, and Meta), which are NVIDIA's largest customers for AI chips, collectively announced roughly $690 billion in AI-related capital expenditures (CapEx) planned for 2026.
Despite many of them beating earnings expectations recently, the market reacted negatively — wiping out about $1 trillion in combined market capitalization in a short period (described as "one week"). CapEx breakdown (approximate figures cited):
Alphabet: ~$180B (doubled from prior year)
Amazon: ~$200B (up ~50% YoY)
Microsoft: $120B+
Meta: $115–135B
(Plus mentions of Oracle at ~$50B in related spending) Wall Street's skepticism: Investors appear unconvinced that this massive AI infrastructure spend will actually generate proportional returns/monetization soon. Stocks dropped sharply (e.g., Alphabet -9%, Microsoft significantly off highs).
NVIDIA's position: The company is portrayed as the big winner here — they reported $57 billion in quarterly revenue (with data center/AI chips at ~$50B, up 66% YoY), and are guiding even higher (~$65B) for the quarter ending February 25, 2026. The video suggests NVIDIA profits massively from selling the picks and shovels (chips/GPUs) regardless of whether the AI "gold rush" ultimately pays off for the buyers.
Bubble warning: The presenter references Fidelity's "5 bubble indicators," claiming the current AI hype meets all five:
Elevated valuations Rapid/unprecedented spending Circular financing (money cycling in ways that inflate hype) Reliance on debt/leverage Lagging or minimal monetization (noted: only ~4 cents in revenue per $1 spent on AI infrastructure so far)
The tone is skeptical/critical of the AI investment frenzy, predicting an eventual correction or "bubble" pop, while highlighting how NVIDIA benefits in the meantime.
(Excerpt) Read more at youtu.be ...
|
|
Click here: to donate by Credit Card Or here: to donate by PayPal Or by mail to: Free Republic, LLC - PO Box 9771 - Fresno, CA 93794 Thank you very much and God bless you. |
MSFT
Still up for the Year.
A five year double.
“The video suggests NVIDIA profits massively from selling the picks and shovels (chips/GPUs) regardless of whether the AI “gold rush” ultimately pays off for the buyers.”
Applied Materials is selling the picks and shovels. Nvidia is selling the ore crushers, hard rock drills and lifting equipment.
Pop goes the weasel.
How about apple?!
If the latest AI models are as good as they say, they shouldn’t need as much need processing power to so the same or better work.
“How about apple?!”
Almost a five year double.
“Reliance on debt/leverage”
That’s not really accurate.
Over 90% of this planned spend is in the bank and belongs to the big hyperscalers.
Free cash flow minus all costs.
They are spending the money they earned from nearly everyone in the world to take the world to a new technological basis.
But nobody is quite sure what that will be.
“They are spending the money they earned from nearly everyone in the world to take the world to a new technological basis.”
Yup.
CATerpillar.
A one year triple
Caterpillar Unveils AI-Powered Future and Invests in the Workforce Building It
Caterpillar.com
AI…. usually ‘the juice is nor worth the squeeze’. The financial returns for the hundreds of billions invested make it frequently NOT FINANCIALLY FEASIBLE.
One organic human brain = about 90 million neurons = 100-500 trillion connections. Grok 4 is about 3 trillion parameters (not that they’re 1-1, but hey), costs $490M to train, and took 200k GPUs. So, playing very fast and loose with numbers one could say they’re about 100x behind anything like a human-equivalent AI.
We do this work on 20w power. Grok took 310 gigawatt-hours to train.
If this thing isn’t going to collapse itself the designers need to be thinking energy efficiency. They’ll hit that wall before they get to where they want to be.
Source:
https://epoch.ai/data-insights/grok-4-training-resources
I have seen a few of this guys videos.
I want the bubble to hurry up a burst so I can upgrade my computers with all that cheap stuff ,LOL
Yeah, but compared to the handwringing here of NVDA which is up 10x over the past 5 years I feel that should get some inclusion here for perspective. Especially as Apple appears to have dropped the ball with AI (though I contend that might be a good thing in the long run as they might be able to introduce a more “seasoned” solution than what we’re seeing now from all the tech companies.
As a long time engineer it’s damn scary.
A year ago coding with it was a joke. In the past few months they’ve introduced new models which range from still a joke to “I need this code ported from Python to Java and it has to also convert the Python specific library calls to the Java equivalents already used in the project (they weren’t the common choices.)
AI spit out the module in under 5 minutes and it’s 95% correct. (There was some special code handling to get around some Python issues that wasn’t needed here)
I have to think part of the issue for Meta and Alphabet is the class action lawsuit in Cali regarding the addictive nature of social media. There is no way a California jury isn’t out to heavily pick some pockets.
Exactly! The delta in power consumption is all you need to see to know that the arch is wrong. Also, it ignores quantum effects.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.