AI Verse β tracking one story: how AI is reshaping infrastructure, stocks, and crypto.
π§ Chips, cloud, energy, data centers
π AI-driven stocks, analyzed not hyped
πͺ Crypto where it intersects with AI
Daily. Not financial advice. This is day one.
π¨ 4 HIDDEN STOCKS BEHIND THE AI BOOM β EP.06
Everyone sees the GPUs. But AI needs an entire physical machine around them:
β‘ $AGX β Generate the power
π $FPS β Distribute the power
π $SMTC β Move the data
βοΈ $WOR β Remove the heat
AI isnβt just a chip story. Itβs an infrastructure story.
@DefiDragans@YouTube Fair point, but low orbit is only ~500 km up so the light delay is a few milliseconds. I think cooling (radiators only, no air) and launch cost are the bigger problems. The 2027 laser-link test will be interesting.
π SPACEX STOCK JUMPS 7% AFTER 3 LAUNCHES IN 13 HOURS
SpaceX pulled off an extraordinary tripleheader:
π¨βπ NASA Crew-13 β ISS
π°οΈ Transporter-18 β 130 payloads, including Googleβs orbital AI prototype
πΊπΈ Falcon Heavy β classified NRO mission Benzinga
$SPCX surged 7%+ near $159 today.
ROCKETS β STARLINK β NATIONAL SECURITY β ORBITAL AI
SpaceX is becoming much more than a launch company.
@DefiDragans The striking part isnβt just the drones β itβs autonomous warfare becoming important enough to justify its own four-star command. Experiments β doctrine β infrastructure. Thatβs a major shift.
π¨ THE PENTAGON WANTS A 4-STAR COMMAND FOR AUTONOMOUS WARFARE.
The proposed AutoWarCom would scale drones, robotics and AI-enabled autonomous systems across the U.S. military β with dedicated budgets, acquisition authority and career paths. (:-Reuters)
Target: October 2027, subject to congressional authorization.
Drones β Robotics β AI β Autonomous Warfare
This is no longer just military experimentation. Itβs becoming military infrastructure.
π¨ TRUMP ADMINISTRATION REPORTEDLY PLANS A $4B NUCLEAR LOAN FOR VISTRA
The financing would help upgrade 3 Vistra nuclear plants as U.S. electricity demand surges β increasingly driven by power-hungry AI data centers. TradingView
Vistra already has major nuclear power agreements with Amazon + Meta. Vistra Corp. Investor Relations
AI DATA CENTERS β POWER DEMAND β NUCLEAR EXPANSION β $VST
The AI infrastructure race is becoming an energy race.
π¨ ANTHROPIC WANTS TO TRAIN 10,000 βFRONTIER DEPLOYED ENGINEERSβ
#Anthropic is investing $100M in its new Claude Frontier Academy.
Engineers from Morgan Stanley, Deloitte, McKinsey, Accenture + others will train, then spend 12 weeks deploying real Claude projects inside their organizations.
The AI race is shifting:
Build the models β train the workforce β deploy AI everywhere.
@DefiDragans Pretty much the direction itβs heading β except instead of mortgages backing the financing, itβs billions of dollars of GPUs generating AI compute. If these structures scale, silicon could become a whole new class of financeable infrastructure assets. Wild to watch.
π¨ AMAZON WANTS WALL STREET TO FINANCE $8B OF NVIDIA CHIPS
Amazon is reportedly exploring moving ~$8B of Grace Blackwell GPUs into an investor-backed vehicle β then leasing the chips back.
Think about that:
Investors β finance GPUs β Amazon leases them β GPUs generate AI compute
AI chips aren't just infrastructure anymore.
They're becoming financial assets.
$AMZN $NVDA
This could become one of the biggest infrastructure questions of the AI-agent era.
If billions of autonomous agents eventually buy data, compute and services from each other, they'll need a financial layer built for machines.
AI agents may need their own economy. π
π¨ NVIDIA IS KNOCKING ON THE DOOR OF A NEW HIGH.
$NVDA is testing the ~$236 area as the catalysts stack up:
β’ Morgan Stanley: Top Pick
β’ $300 price target
β’ $235B buyback capacity
β’ AI demand still accelerating
β’ Treasury yields falling today
$236β237 is the level to watch.
A clean breakout could put NVIDIA back into price discovery. π
π¨ GPT-6 ASTRA CRACKED A 217-YEAR-OLD NAPOLEONIC CIPHER.
Starting with a single scan, Astra transcribed ~1,300 encrypted symbols, found a partial key, built a solver and decoded the rest β in roughly 6 hours of model time.
It even helped correct the letterβs date from 1807 β 1809 by cross-checking historical records.
Vision β Research β Coding β Cryptanalysis
AI just unlocked a message hidden for more than two centuries.
π¨ AI IS MOVING DEEPER INTO CYBERSECURITY.
Googleβs Gemini 4 Argon can autonomously find, validate + patch vulnerabilities, while OpenAI is pushing Astra deeper into cyber.
But cyber stocks arenβt broadly selling off today β $PANW , $CRWD , $FTNT and $ZS are actually higher.
The bigger question:
Will frontier AI disrupt cybersecurity companies β or make their platforms far more powerful? π
π¨ TRUMP: U.S. COULD TAKE STAKES IN AI GIANTS
Asked whether the U.S. government could take equity in OpenAI, Anthropic and other frontier AI companies β similar to its Intel stake β Trump said:
βI might. Maybe I could do that.β
Not a confirmed policy yet.
But the direction is striking:
Regulate AI β Fund AI β Own part of AI
The U.S. government could eventually become a shareholder in the frontier AI race.
π¨ AI IS BOOMING β BUT 5.3% TREASURY YIELDS ARE FIGHTING THE RALLY.
Micron crushed earnings. AI demand remains huge.
Yet $MU fell ~2% as the U.S. 10Y briefly hit 5.34% β the highest since 2002.
AI earnings β
Cost of money β
Valuations under pressure
The AI trade isnβt dead. Money is getting expensive.
π¨ MICRON CRUSHED EARNINGS β BUT THE STOCK BARELY MOVED.
$MU Q4 revenue hit $54.23B, adjusted EPS $33.42, while Q1 revenue guidance came in at $61.5B Β± $1.5B.
AI-driven HBM and data-center demand remain exceptionally strong.
Yet $MU barely reacted after hours.
Why? Expectations were already sky-high after the stockβs massive run.
AI demand is booming β but at these valuations, even blowout earnings may not be enough.
π¨ GOOGLE UNVEILS GEMINI 4 ARGON
Googleβs new frontier model leads 12 of 16 benchmarks in its published comparison against GPT-6 Astra and Claude.
Built for complex workflows, coding + cyber defense.
But Google isnβt broadly releasing it yet.
Capability β β caution β
Lots of discussion out there about our next model(!), so I wanted to give an early look as soon as possible. Introducing Gemini 4 Argon!
It shows frontier performance in complex workflows, cyber defense and software engineering. Teams are using it extensively at Google, from coding to quantum computing, great feedback.
Hereβs a look at the benchmarks:
π¨ AMAZON JUST SIGNED A $1B+ DEAL WITH SYNOPSYS.
AWS is expanding its use of $SNPS chip IP as Amazon pushes deeper into custom silicon β including Trainium AI accelerators and Graviton CPUs.
AWS β CUSTOM CHIPS β AI COMPUTE
$AMZN $SNPS