| VC Tech Bro
| Forbes 30u30
| AI Doomer Dude
| Gen Z Threadboi
| King of Hot Takes
| 27 y.o w/ $500k NW
| Chase Coleman's Nephew
| CSO (Chief Shitpost Officer)
Sep 3 • 20 tweets • 6 min read
How to find 10 baggers according to Mark Mahaneyđź§µ
10 timeless lessons for public tech investors.
The dude has covered tech stocks for the last 25 years
Aug 28 • 12 tweets • 3 min read
$40k -> $600
Unreal.
Jesus fucking Christ man.
Aug 16 • 9 tweets • 3 min read
đź§µWhy the mag 7 are about to implode:
1/ The AI capex wave by $AMZN $MSFT $GOOG $META due to rising gen AI inference workloads has a significant unintended consequence the market is not ready for.
The hidden AI killer is IMPAIRMENT LOSS (led by overusage and tech obsolescence). 2/ In 2025, AI capex spend will hit >$300b, funneled toward data centers & GPUs to support generative AI workloads.
This is rational given training and inference demand, but it overlooks a critical issue: hardware is being utilized at rates FAR exceeding original design specs.
Aug 16 • 11 tweets • 5 min read
The "AI Trade" has largely played out in the US in the last 3Y since ChatGPT led to:
An explosion in:
GPU spend $NVDA
Hyperscaler inference $MSFT $GOOG $AMZN
AI startups founded
Missed the US AI trade? This is how you can catch the China AI trade (which hasn't even begun yet):
1. The bottom of the stack: GPUs.
Personally, I think this stack is not as investible. Just long $NVDA as most China cos will buy H20s.
China doesn't have a "Dominant GPU champion" that you can buy given Huawei is private & the chips are inferior.
$SMICY $HHUSF are far behind
Aug 14 • 17 tweets • 4 min read
Why AI is a house of cards:
1. You pay $200 a year for an AI app (like Cursor).
2. Cursor pays OpenAI $500 for API tokens ($300 of which is VC funding).
3. OpenAI pays AWS $1000 for compute ($500 of which is VC funding).
4. $AWS pays $10k for $nvda GPUs.
See the problem?
Unless you as a user are miraculously comfortable paying $1k for the AI app
The only thing propping up AI is VC funding
No VC funding:
The AI application layer is unprofitable
The LLM layer is unprofitable
The compute layer is unprofitable
The GPu layer is unsustainable
Aug 5 • 11 tweets • 3 min read
Your job is to find a way to hit $1-3 million and become unemployed by age 40.
Make your money
This strategy becomes increasingly more attractive as interest rates get closer to 0%, margin rates drop, CD/risk free rate drops, discount rate drops and equities rip higher
Jul 19 • 4 tweets • 1 min read
$BIDU has a current market cap of 30 billion.
$BIDU has net cash + investments of $28 billion.
In other words, you effectively get: 1. $BIDU core search biz 2. Apollo self-driving biz 3. $BIDU cloud
All for free. All upside. This might be the most asymmetric opportunity ever
Let's say hypothetically the three biz's went to $0 overnight and they roll up the entire company tomorrow.
- They sell all the stakes in public companies @ market value over 6 months.
- They distribute net cash back to shareholders.
You effectively get your $$ back 1:1.
Jul 5 • 6 tweets • 2 min read
$PDD De minimis removal basically vaporized the fully managed model GMV
$TEMU still cheaper than $AMZN even after factorign in De Minimis