AI might really change the world, but most AI companies probably won't make it to that day.
Right now, the most dangerous thing in the entire market is:
It's not that people don't believe in AI.
It's that everyone believes in AI too much.
They've even started to front-run the growth of the next 10 years,
by pushing it all into today's stock prices.
Many AI companies are currently going all out with crazy funding, crazy expansions, crazy GPU purchases, and crazy data center builds.
But the problem is:
Most of them aren't even making a profit yet.
So why are they still burning cash like crazy?
Because everyone is scared:
"If I don't expand today, I'll be out of the game tomorrow."
Doesn't this sound like the 2000 internet bubble?
Back then, all companies were also rushing to lay fiber optics, build data rooms, and grab users.
In the end, the internet did change the world.
But in the process:
A lot of companies went belly up.
The ones that actually survived were only a handful of winners like Amazon and Google.
Now, AI is pretty much the same.
🚨The tech revolution is real.
🚨Industry demand is also real.
But market valuations aren't necessarily real.
And many people haven't noticed:
The biggest bottleneck in AI
is no longer the models.
But rather:
⚡ Power
⚡ GPU capacity
⚡ Data centers
⚡ Optical modules
⚡ Storage chips
⚡ Network bandwidth
⚡ Cooling systems
The entire AI industry is essentially still in the:
"Building the highways first" phase.
The truly sustainable revenue models are still maturing.
So right now, the most profitable ones
are actually not the model companies.
But the ones selling the picks and shovels.
NVDA, optical modules, HBM, data centers, power,
these are the real big winners in this AI wave.
Because during the gold rush:
The ones who consistently make money
were never the gold miners.
But the ones selling the picks and shovels.
One last thing:
The AI bubble may not destroy AI.
But it will definitely wipe out a lot of follower companies.
#Ai $NVDA