@tkspears2 Betting on robots, nuclear, and space so I can retire on Mars while my humanoid robot waters the plants back home 😂
In all seriousness, this convergence thesis is extremely strong. $OKLO + $ASTS combo has me hyped.
Wealth Unlocked by 2035.
My 10 most promising plays from the labs, launchpads, and data centers of tomorrow:
1. $QCOM — Edge AI chip leader powering billions of connected devices.
2. $ORCL — Enterprise cloud + AI database infrastructure giant.
3. $ETN — Power management & electrical infrastructure for AI data centers.
4. $NVT — Critical cooling & electrification solutions for hyperscale facilities.
5. $KTOS — Autonomous systems & defense AI technology leader.
6. $TTD — AI-powered digital advertising & marketing platform.
7. $SMR — Small modular nuclear reactors built for AI power demand.
8. $GTLB — AI-enhanced DevOps and software development platform.
9. $RBRK — Modern data security & backup for the AI era.
10. $TEAM — AI-first collaboration & productivity software suite.
Innovation is rarely a straight line.
But these tracks look built to last.
If you had to drop one and double down on another, which would it be?
People think they missed the memory trade.
I don’t think they have even seen the middle of it.
Right before the AI boom hit $MU, $SNDK
, and $SK Hynix all cut production. The memory cycle had been brutal. Prices were collapsing. The rational move was to slow output and wait for demand to recover.
Then ChatGPT happened. Then agentic AI happened. Then every hyperscaler on earth started building data centers at a scale nobody had ever planned for.
Memory demand exploded overnight and the companies that make it had just deliberately slowed their production lines.
You cannot turn a semiconductor fab back on like a light switch. These are multi-year buildouts. The fabs being built today do not reach full production until 2028 to 2030.
That is the structural story. A demand curve that compounds every year with AI and a supply response that is years behind it.
Now look at the valuations.
$MU trades at roughly 10x forward earnings despite 196% revenue growth and 75% gross margins.
$SNDK is still cheap relative to its locked in revenue commitments.
$SK Hynix is printing $26 billion in operating profit per quarter on an HBM monopoly and still trades at a discount to US peers.
The forward multiples on these companies do not reflect a shortage that runs through 2029.
$DRAM is the cleanest way to own all of it. One ticker. Every single company within wins from the same structural imbalance.
The shortage created the trade. The valuations keep it going.
@PaulKeitzke Intel at $300? Let me cry for a second Everything else I can believe, but INTC still hurts. Overall semiconductor bull thesis is rock solid though.
There’s now an overwhelming general consensus on Wall Street that we are going to see 8,000 on the $SPX in the next 12 months or maybe even way sooner.
Comment + Repost this post for a chance to win a $20 stock!