Everyone spent this week refreshing the ai price war... gpt 5.6 at $1, grok 4.5, muse spark 1.1 meanwhile, in a corner nobody was watching, china quietly caught a 63m rocket at sea, on its maiden flight:

-> took 5 barge attempts over 15 months...
-> lm-10b did it on try one
zero landing legs: a cross-shaped arresting net + one hook on the booster, an architecture nobody has flown before

It’s huge

💥 IRAN WAR - THE MOVIE💥

#Oil just got a fresh bid.

Trump and Iran as soon as the stock market closed

Three companies still control ~90% of the global DRAM market:

• Samsung ~37%
• $SKHY ~30%
• $MU ~23%

The important point is that the value inside DRAM is shifting toward HBM and server memory where the big three have the scale, process know-how and customer relationships to capture the highest-margin demand.

So even as CXMT competes more aggressively in commodity DDR5 and LPDDR, the oligopoly at the top of memory stack may actually be strengthening with Samsung, SK Hynix and Micron concentrating capacity on the scarce, high-margin parts of the market.

Also important to call out CXMT is still operating under sanctions with restricted access to key $ASML and $AMAT equipment which makes its path to leading-edge DRAM and HBM much harder.

After the game, the king and the pawn go into the same box

🇺🇸 Trump highlights $MU Micron's big move: The company is accelerating its U.S. investment to $250B through 2035 to build advanced memory chips domestically.

This ramps up previous plans and is expected to create thousands of manufacturing jobs while strengthening American supply chains for AI and tech.

Trump credits deregulation and a pro manufacturing push for the momentum.

What money actually buys:

$10k = a safety net
$100k = options in life
$1m = freedom from worry
$5m = full financial freedom
$10m = time becomes yours
$50m = access to anything
$100m = problems disappear
$500m = you buy entire companies
$1b = you enter a different world
$10b = governments take your calls
$50b = you fund entire nations
$100b = you start shaping the world

$HIMS Dip Buying Opportunity in Telehealth Leader

The Pure Technicals:
✅ Holding critical weekly support with higher low structure forming
✅ Fonsie Channels coiling for potential bullish breakout
✅ Volume profile supports accumulation on this pullback
✅ Clear technical risk/reward to the upside

Key Levels: Near-term Target: $40 – $46 Stretch Target: $55+ Support: $31.85 – $34

Quick DD:

Dominant personalized health platform with GLP-1 momentum and expanding offerings

International growth + AI investments fueling next leg of scale

Strong subscriber economics in massive consumer healthcare TAM

Volatility creates attractive entry for long-term conviction

Patient capital wins in names like this.

$AIBZ Showing Heavy Accumulation – AI Data Center Infrastructure Name Gaining Traction

Bitzero Holdings ($AIBZ) is catching attention with strong volume accumulation on the daily chart. The stock is up +8.48% today near $7.80, with a clear box of elevated green volume during the recent consolidation phase — classic institutional footprint behavior after the initial Nasdaq listing momentum.
Why this fits the ISAR thesis:

Provider of IT energy infrastructure and high-efficiency power for data centers.
Focus on HPC (high-performance computing) and AI workloads alongside Bitcoin mining.
Operations in North America and Nordic regions with access to clean, low-carbon energy sources.

This directly addresses the AI Infrastructure pillar — where power delivery and data center capacity remain the primary bottlenecks for scaling. As AI demand surges, names that can deliver sustainable energy infrastructure solutions are strategically positioned to benefit, feeding into the broader convergence toward Robotics and Space applications.

The chart shows a healthy base-building pattern post-rally, with price holding key support levels and volume confirming buyer interest on dips. Continued accumulation here could set up the next leg higher if AI/power themes stay in favor.

Key watch:
Sustained volume and ability to hold above $7–$7.50 zone would strengthen the setup.

Another emerging name in the AI Energy Infrastructure space worth monitoring as the macro narrative plays out. Energy enables AI. Infrastructure scales it. $AIBZ is one to keep on the radar.

Position sizing and risk management essential on these volatile infrastructure plays.

$FCEL Pullback – Potential Final Opportunity at Discounted Levels?

FuelCell Energy is pulling back sharply today (-8.57% to ~$21.03) amid broader market action, but the longer-term daily structure remains constructive. Price is testing the upper trendline of the prior base while holding above key moving averages and within the overall bullish channel established since the spring breakout.
This move creates an interesting risk/reward setup for those aligned with the ISAR thesis. With recent catalysts (Siemens collaboration, $225M capital raise, and data center pipeline momentum), the dip may represent one of the last attractive entry points before potential reacceleration driven by AI power demand.
Technical context:

Testing trendline support from the May-June rally
Elevated volume on the way up earlier; current volume is elevated on the pullback
Broader channel points to significant upside if support holds

Green hydrogen and fuel cell infrastructure sit at the heart of the Energy bottleneck for AI scaling. As data centers and industrial electrification accelerate, $FCEL’s ability to deliver scalable on-site power becomes increasingly strategic — directly enabling the convergence toward Robotics and Space applications.

Watch levels:
Holding above $20–$21 zone with volume stabilization would be constructive. Failure to defend could test lower channel support, but the fundamental setup tied to AI energy needs keeps the longer-term bias bullish.
High-conviction names in the infrastructure layer often offer these volatility-driven opportunities. For believers in the macro narrative — AI needs energy, energy needs infrastructure — this could be a tactical entry window.

$CTNX Bottom Breakout Watch in Biotech
🔬
The Pure Technicals:
✅ Strong green candle reclaiming key micro-support with volume conviction
✅ Fonsie Channels beginning to tilt higher off multi-month lows
✅ Clear higher low formation in the base
✅ Oversold bounce with room for continuation

Key Levels: Near-term Target: $0.85 – $1.10 Stretch Target: $1.83+ Support: $0.62 – $0.70

Quick DD:

- Clinical-stage biotech advancing novel therapies in oncology with focused pipeline

- Low-float setup capable of rapid moves on positive clinical or partnership catalysts

- Recent volume surge signaling renewed interest at these depressed levels

- Asymmetric risk/reward typical of small-cap biotech bottoms

Watching closely for confirmation of the bottom break.

$TRAW +12.6% Bounce in Biotech Antiviral Play 📈

The Pure Technicals:
✅ Strong daily reversal candle holding key support zone
✅ Fonsie Channels showing compression ready for expansion
✅ Volume confirming buyer defense at multi-week lows
✅ Bullish structure forming above $0.65 base

Key Levels: Near-term Target: $0.94 – $1.15 Stretch Target: $1.53+ Support: $0.69 – $0.76

Quick DD:
Clinical-stage biotech advancing novel oral antivirals targeting influenza, COVID-19, bird flu, and emerging threats like Hantavirus

Pipeline includes Tivoxavir marboxil (influenza) and Ratutrelvir programs with recent regulatory and toxicology updates

Low-float biotech with high short interest (7.6% of float) sets up for rapid moves on positive pipeline news

Positioned in critical respiratory antiviral space with multiple shots on goal

Resilient bounce in a name with real clinical catalysts ahead

$YAAS Surged 35% this week following the TikTok ecosystem expansion catalyst.

The Pure Technicals:

✅ Massive volume surge confirming institutional accumulation on the weekly breakout
✅ Strong rejection from multi-month lows with decisive green candle structure
✅ Fonsie Channels showing early bullish expansion after prolonged compression phase
✅ Clear higher high formation with momentum building toward overhead resistance

Key Levels: Near-term Target: $1.80 – $2.57 Stretch Target: $3.79+ Support: $0.72 – $0.93

Quick DD:
Strategic 18% stake acquisition in YATOP (valued at ~$60M) brings Tier-1 TikTok agency capabilities, 10k+ creator network, and teams in LA/Hong Kong
Integrates directly with Youxin's SaaS/PaaS platform to accelerate global digital commerce and marketing services
Low float (~6.7M) with minimal institutional ownership sets stage for explosive moves on continued momentum

Building on prior Celnet acquisition to strengthen AI/CRM offerings in high-growth digital sectors
This is exactly the type of asymmetric setup where smart capital rotates early into undervalued tech plays with real ecosystem expansion.

FYI the Q2 earnings season kicks off next week

Top 20 mining firms by revenue

Everyone ranks mining by revenue and assumes gold dominates.

39 of the top 100 mining companies by headcount are gold miners.

They generate 21% of industry revenue.

22 copper producers generate 28%, over $200 billion, more than any other metal.

That is why Rio Tinto is circling Glencore. Its copper business is what pushed Glencore to $63 billion in revenue, the highest in the sector.

Copper is the real cash engine of mining

If you still doubt it, here are 16 humanoid robot hands moving closer to the shape, reach and dexterity of a human hand.

01 🇳🇴 1X NEO: 25 DoF, tendon drive, 3-DoF wrist, pressure, location and slip sensing, up to 45 N finger force, 17.75 Nm wrist torque and IP68 protection.

02 🇺🇸 Tesla Optimus: Human-scale architecture integrating compact actuators, sensors and mechanical transmissions. Tesla has not released a complete public hand specification.

03 🇺🇸 Figure 03: Soft adaptive fingertips, tactile sensors and palm cameras designed to stabilize grasps across different objects.

04 🇨🇦 Sanctuary AI Phoenix: 21-DoF hydraulic hands with tactile sensing and demonstrated in-hand object manipulation.

05 🇺🇸 Apptronik Apollo: Designed for dexterous manipulation. Apptronik has said that more human-like hands remain part of its continuing R&D work.

06 🇨🇳 Unitree G1: Optional force-controlled hands. Dex3-1 provides 7 DoF, while the Inspire option uses 6 DoF across 12 moving joints.

07 🇨🇳 Fourier GR-1: Human-like hand geometry for versatile grasping. The newer GR-2 advances the platform to 12-DoF hands with six tactile sensor arrays.

08 🇨🇳 UBTECH Walker S: Five-finger dexterous hands with pressure sensing. Walker S2’s fourth-generation hand adds an active thumb DoF and raises fingertip payload to 12.5 N.

09 🇨🇳 Robot Era STAR1: Fully actuated dexterous hands developed for precise object manipulation.

10 🇨🇳 EngineAI PM01: A compact humanoid platform with five-finger hands, although detailed official hand specifications remain limited.

11 🇨🇳 AgiBot A2: The A2 family includes industrial hands with 19 DoF, 12 actively driven. AgiBot’s OmniHand offers 16 DoF and more than 400 force-sensing taxels.

12 🇨🇳 XPENG IRON: Human-scale 1:1 hands with 22 DoF, built around XPENG’s smallest harmonic joints.

13 🇨🇳 Kepler Forerunner K2: Developed for industrial manipulation and object handling. Detailed hand-level specifications have not been published.

14 🇨🇳 Xiaomi CyberOne: Five-finger humanoid hands publicly demonstrated for object handling, with limited published hand-level specifications.

15 🇯🇵 Toyota T-HR3: Ten articulated fingers, torque sensing and force-feedback teleoperation through Toyota’s Master Maneuvering System.

16 🇪🇸 PAL Robotics TALOS: Torque-controlled manipulation for grasping and operating tools. Its 7-DoF arms can carry 6 kg each with the arm extended.

These hands are not equally mature. Some are commercial products, some are optional end-effectors and others remain research prototypes.

What task do you believe is still impossible for a robot?

The AI models that will generate the most tokens will be different than the models that generate the highest revenue

The way things are headed, it is very clear that low value repetitive tasks will be outsourced to the cheap and open source models

But the cutting edge work done with AI will consolidate around 1 or 2 frontier models, likely led by Anthropic and OpenAI

So while the cheap models will produce the most tokens, it will not provide nearly the same value. This will be the commodity layer of models

The frontier LLMs will be low token generation but high value. This will be the premium layer.

This is why we invest in #Bitcoin.