**Summary:**

NBIS confirmed its $230 breakout (tagged $249.19), CoreWeave's JPMorgan upgrade drove it to $91.12 before Friday's fade, and IREN broke down below $44. This week's AI Infrastructure Weekly grades last week's levels against Friday's closing charts — no softening missed calls.

NBIS closed at $237.33 (+6.2% on the week) after BNP Paribas upgraded to Outperform with a $399 target and Bank of America raised FY2027/FY2028 revenue forecasts. CoreWeave closed at $87.59 after JPMorgan's upgrade to Overweight ($125 target) sparked a run to $91.12 that couldn't hold. IREN closed at $44.13, below its 200-day average, after Rothschild Redburn initiated at Neutral with a $40 target — the $45-46 support and $44 invalidation line both failed.

Summary

AI infrastructure is one of the most exciting themes in the market — and one of the easiest places to get hurt if you’re trading the story instead of the setup.

In this episode, we revisit three high-volatility AI infrastructure names — IREN, Nebius (NBIS), and CoreWeave (CRWV) — and examine what happened after we established specific technical levels for each stock.

The result is a surprisingly clean lesson in three core trading concepts:

IREN = Confirmation.

The stock broke through the $44.50–$45 confirmation zone and advanced toward $46–$47. Momentum improved, but the trade has now changed: rather than chasing, we’re watching $48–$49 for continuation or a controlled retest of the prior breakout zone.

NBIS = Patience.

The crucial $220–$221 support area survived, but the stock still hasn’t reclaimed the $230 confirmation level. The higher-time-frame trend remains constructive, but the framework says to wait for the market to provide evidence before committing.

CRWV = Invalidation.

CoreWeave needed to hold $88. It didn’t. The stock subsequently fell toward the $80–$81 area while moving below its major daily moving averages. The lesson isn’t that CoreWeave suddenly became a bad company; it’s that the specific swing-trade thesis failed.

The larger point is simple:

A great company is not automatically a great trade.

Technical levels give investors a framework for determining in advance what would strengthen a thesis, what requires patience, and what proves the trade wrong — before fear, greed, or FOMO take over.

We also explore the idea of “paying for information”: deliberately accepting a slightly higher entry price in exchange for additional confirmation that institutional buyers are actually stepping in.

And at the end, we take the concept beyond investing: could defining our own personal “invalidation levels” before becoming emotionally committed help us make better decisions in careers, projects, and relationships?

Trade the levels, not the hype.

This episode is for educational and informational purposes only and is not individualized financial advice.

Summary:

AI infrastructure stocks are moving fast — but which one offers the best swing-trade setup right now?

In this episode, we break down three of the most volatile names powering the AI infrastructure boom: Nebius (NBIS), IREN (IREN), and CoreWeave (CRWV).

Rather than focusing on the 10-year AI thesis, we’re looking specifically at the technical setups heading into the week of September 14, 2026 — including key support, resistance, EMA levels, breakout triggers, downside invalidation levels, and potential price targets.

🏆 #1 NBIS — Swing Score: 8.1/10

The strongest technical setup of the three. NBIS remains in a powerful uptrend, with the 20-day EMA above the 50-day and both well above the 200-day EMA. The key battleground is $220–$221, while a reclaim of $230 could confirm renewed momentum toward $235 and potentially $245–$250.

🥈 #2 IREN — Swing Score: 7.6/10

IREN is sitting above a tight cluster of major moving averages around $42. A move through $44.50–$45 could confirm another leg higher toward $46, $48–$49 and potentially $50+. But losing $40 would materially weaken the setup.

🥉 #3 CRWV — Swing Score: 6.8/10

CoreWeave remains trapped in a difficult compression zone around its short- and intermediate-term moving averages. A breakout through $91–$92 could open a move toward the 200-day EMA around $95–$96 and eventually $100+, while losing $88 increases downside risk.

We also examine why Federal Reserve policy, interest rates and financing conditions matter so much for capital-intensive AI infrastructure companies — and why technical analysis is less about predicting the future than establishing your battle lines before volatility arrives.

The biggest lesson: don’t anticipate the breakout. Make the market prove it.

This is not financial advice. This content is for educational and informational purposes only.

Episode Description

IonQ just reported the strongest quarter in its history.

Revenue hit $80.1 million — up 287% year over year — while the company simultaneously reported a staggering $1.87 billion GAAP net loss.

At first glance, those numbers seem impossible to reconcile.

But buried beneath the headline loss is a much more interesting story.

In this episode, we break down IonQ’s Q2 2026 earnings and examine why much of that enormous GAAP loss came from fair-value accounting rather than the operating performance of the underlying business.

Then we go much deeper.

IonQ is attempting to transform itself from a quantum-computing company into something considerably larger: a vertically integrated quantum infrastructure platform spanning computing, semiconductor fabrication, networking, sensing, security and photonics.

We examine:

• Why IonQ’s remaining performance obligations have surged to roughly $485 million
• Why the $1.8 billion SkyWater acquisition could fundamentally alter IonQ’s manufacturing economics
• How Oxford Ionics’ electronic qubit control could move trapped-ion computing away from complex laser laboratories and toward semiconductor-style manufacturing
• Why quantum error correction may represent one of the most important technical milestones in the company’s roadmap
• IonQ’s emerging merchant-supplier strategy
• Why photonics, quantum networking and sensing may become businesses independent of quantum computers themselves
• The importance of IonQ’s 84 optical communications terminals already in orbit
• Why the approaching threat of Q-Day is changing enterprise buying behavior
• How quantum key distribution could become a new cybersecurity infrastructure layer
• And why IonQ may ultimately benefit from both sides of the quantum revolution

Because IonQ is pursuing a fascinating dual strategy.

It is helping build increasingly powerful quantum computers capable of threatening today’s encryption — the lockpick.

At the same time, it is developing quantum networking and security infrastructure designed to protect the world against those machines — the lockbox.

If both markets develop as IonQ expects, the company may be attempting something much more ambitious than simply winning the quantum-computing race.

It may be trying to become the infrastructure layer through which much of the quantum economy operates.

Summary:

SpaceX may look like a rocket company—but this earnings-call deep dive argues that rockets are becoming only the delivery layer for a much larger infrastructure empire.

We break down the extraordinary claims surrounding SpaceX’s Q2 FY2026 performance and the three businesses that could define its future: AI compute infrastructure, Starlink connectivity/mobile, and Starship launch economics.

The central investment thesis is powerful: if SpaceX can combine cheap reusable launch, a massive satellite communications network, and vertically integrated AI infrastructure, it could attack markets far larger than commercial spaceflight itself.

This episode explores:

• Why AI compute could become a major SpaceX business
• How aerospace thermal-engineering expertise could translate into data-center infrastructure
• Why Starlink may be the connective tissue tying SpaceX’s businesses together
• The potential for direct-to-device satellite communications to challenge terrestrial telecom
• Why Starship’s economics matter far beyond Mars
• How dramatically lower launch costs could change the economics of orbital infrastructure
• Why vertical integration is arguably SpaceX’s greatest competitive advantage
• And the provocative possibility that SpaceX eventually becomes less an aerospace company and more a global infrastructure platform

The bigger question isn’t simply whether SpaceX wins in space.

It’s what happens if SpaceX uses its dominance in space to attack some of the largest industries on Earth.

📝 Episode Summary:

Meta’s newest AI strategy may be far more than another Silicon Valley manifesto.

In The Future Is for Everyone, Meta lays out a vision where billions of ordinary people gain access to personal superintelligence capable of coding software, designing products, tutoring children, providing professional expertise, and helping individuals turn ideas into businesses.

But beneath the philosophical language is a potentially enormous business strategy.

In this episode, we break down Meta’s plan from an investor’s perspective: why Meta is betting on personal rather than centralized AI, how superintelligence could finally provide the killer application for Meta’s massive investment in AI glasses and Reality Labs, and how Meta could potentially monetize scarce AI compute through a dynamic auction model.

We also examine the contradictions and risks. If access to greater intelligence ultimately depends on the ability to purchase more compute, does Meta’s supposedly democratized AI ecosystem simply recreate existing inequalities? And is its argument for distributed AI safety partly a regulatory strategy designed to protect its business model?

Finally, we explore a much bigger question: if one person equipped with AI can effectively command the capabilities of developers, marketers, lawyers, researchers, and administrative staff, what happens to traditional venture capital—and where does economic value migrate next?

This isn’t just about Meta building another AI product.

It could be a blueprint for owning the interface, infrastructure, and economics of the next computing platform.

Summary:

Bloom Energy has reached an extraordinary inflection point.

After taking 21 years to generate its first billion-dollar revenue year, the company is now guiding toward nearly doubling annual revenue again in just a single year.

But beneath the spectacular headlines lies the real story: Bloom has quietly engineered one of the strongest operating leverage models in industrial manufacturing. Revenue surged 166% year over year while operating expenses increased only 48%, allowing profitability to expand dramatically.

The episode explores why artificial intelligence has fundamentally changed the economics of electricity. For hyperscale AI companies, every month of delayed power can translate into billions of dollars of lost token revenue. In that environment, speed—not simply low electricity prices—becomes the most valuable product.

We analyze Bloom’s four competitive moats, including permitting advantages, decentralized manufacturing, financing partnerships, and its solid oxide fuel cell technology that bypasses many of the regulatory challenges facing traditional combustion-based power generation.

The discussion also examines Brookfield’s $25 billion financing commitment, Bloom’s remarkable turnaround in service margins, geopolitical concerns surrounding scandium sourcing, and the long-term opportunity presented by distributed AI inference data centers.

Finally, we ask the critical investment question:

Is Bloom Energy experiencing a temporary AI boom—or building the foundational infrastructure platform for the next generation of computing?

Episode Summary

Tesla’s core automotive operation remains the cash-generating foundation of the company, but management appears increasingly willing to sacrifice near-term margins and capital efficiency to fund its AI, autonomy, robotics and energy ambitions.

The investor thesis rests on five interconnected components:

1. Automotive: Generates cash, vehicles and real-world data.
2. FSD and Robotaxis: Convert vehicle hardware into recurring software and transportation revenue.
3. Custom AI Infrastructure: Trains the neural networks required for autonomy and robotics.
4. Optimus: Extends Tesla’s AI architecture from vehicles into generalized physical labor.
5. Energy Storage: Supplies the power-management infrastructure required by AI data centers and the broader Tesla ecosystem.

If these programs reinforce one another as management expects, Tesla may ultimately operate less like an automaker and more like a vertically integrated AI and physical-infrastructure platform.

The opportunity is enormous, but so are the execution, regulatory, manufacturing and capital-allocation risks.

⚡️Summary:

Wall Street is betting billions that Bloom Energy can solve one of the biggest challenges facing artificial intelligence: power. Its solid oxide fuel cells promise to bypass the aging electric grid and provide on-site electricity for next-generation AI data centers.

But is the investment thesis built on solid ground?

In this episode, we take a deep dive into Hunterbrook Media’s investigation into Bloom Energy. We explore the company’s fuel cell technology, China’s dominance of the global scandium market, supply chain risks, manufacturing scalability, circular financing, backlog accounting, and the regulatory hurdles facing major AI infrastructure projects.

Whether you’re a Bloom Energy investor, interested in AI infrastructure, or simply want to understand the intersection of geopolitics, energy, and finance, this episode provides a detailed, data-driven analysis of one of the market’s most debated companies.

This discussion is for educational purposes only and is not investment advice.

Summary:

Artificial General Intelligence may arrive far sooner than most people expect—and if Leopold Aschenbrenner is correct, it could trigger the largest industrial mobilization in modern history.

In this episode, we take a deep dive into Situational Awareness, Aschenbrenner’s influential 2024 essay that argues exponential improvements in computing power, algorithmic efficiency, and AI research are rapidly accelerating the timeline toward AGI.

We explore:

• The concept of “counting OOMs” (orders of magnitude)

• Why AI capability continues to scale exponentially

• How reinforcement learning and synthetic data change the game

• The intelligence explosion hypothesis

• Why trillion-dollar AI clusters may become necessary

• The coming electricity and semiconductor bottlenecks

• The geopolitical competition surrounding AGI

• What superintelligence could mean for humanity

Whether or not every prediction comes true, this conversation explores one of the most important technological and economic theses of our time.

This episode discusses the ideas presented in Leopold Aschenbrenner’s essay and distinguishes between documented developments and speculative forecasts.

Summary

Everyone loves TPS numbers, but few understand the architecture underneath.

In this deep dive, we examine how FastPay, Narwhal, Tusk, and SUI separate transaction dissemination from consensus, enable parallel execution, and solve blockchain scalability without relying on hidden settlement assumptions like traditional payment networks.

More importantly, we explore the economic design that makes SUI potentially unique: a storage fund that compensates future validators while simultaneously creating a quasi-deflationary sink for the token supply.

Rather than focusing on hype, we ask a deeper question:

Could SUI ultimately derive most of its value not from transaction speed, but from becoming the most capital-efficient decentralized data layer on the internet?

Key Takeaways

✅ Why Visa’s speed is largely built on promises rather than immediate settlement.

✅ How FastPay achieved 160,000 TPS with finality.

✅ Why Narwhal and Tusk separate data dissemination from consensus.

✅ What “embarrassingly parallelizable” means and why investors should care.

✅ How SUI’s object-centric architecture unlocks parallel execution.

✅ Why shared objects and owned objects matter.

✅ The economics behind state bloat and why most blockchains ignore it.

✅ How SUI’s storage fund solves long-term validator incentives.

✅ Why the tokenomics create a quasi-deflationary mechanism.

✅ How validator game theory prevents fee monopolies.

✅ Why decentralization incentives matter as much as raw throughput.

✅ The ultimate question: Is SUI really a decentralized data layer masquerading as a blockchain?

🚀 Summary:

In this episode, we break down one of the most ambitious business and investment theses ever proposed: SpaceX’s vision for the future of artificial intelligence, telecommunications, and space infrastructure.

What begins as a discussion about rockets quickly transforms into something much bigger. SpaceX is no longer just a launch company. According to its IPO filing, it has become a vertically integrated technology giant spanning three businesses: space transportation, Starlink connectivity, and artificial intelligence through xAI.

We explore how reusable rockets dramatically lowered the cost of reaching orbit, enabling Starlink to become one of the world’s largest satellite internet providers. With over 10 million subscribers and billions in annual profits, Starlink has become the financial engine powering SpaceX’s next phase of growth.

That next phase is AI.

The company argues that Earth’s electrical grids may soon become the limiting factor for AI development. Their solution? Move massive data centers into orbit where solar energy is abundant and the vacuum of space provides natural cooling. Combined with Starlink’s communication network, proprietary AI chips, and fully reusable Starship rockets, SpaceX believes it can dramatically reduce the cost of computing and build the infrastructure for a future space-based economy.

Along the way, we examine the enormous risks: Starship execution, regulatory challenges, orbital debris, massive capital expenditures, and Elon Musk’s complete voting control over the company.

Ultimately, this isn’t just a conversation about a company. It’s a discussion about whether humanity is approaching a future where economic growth extends beyond Earth itself.

📈Summary:

Nebius Group is making one of the boldest bets in AI infrastructure today: spending up to $25 billion despite generating only about $400 million in quarterly revenue. In this episode, we examine how the company transformed into an AI-native hyperscaler, why its cloud business is growing over 800% year-over-year, and how partnerships with ** Microsoft⁠, ** Meta⁠, and ** NVIDIA⁠** are helping fund its massive expansion.

We also explore the risks: execution challenges, data center buildouts, community opposition, and the possibility that future AI breakthroughs reduce the need for massive computing infrastructure. The ultimate question: in the AI gold rush, will the biggest winners be the companies creating the algorithms—or the ones controlling the power and compute behind them?

🥗Summary:

In this deep dive, we unpack one of the most fascinating paradoxes in modern fast-casual investing: why did Sweetgreen spend years and millions building revolutionary kitchen automation technology… only to sell the entire robotics division just as it started working?

Using Sweetgreen’s Q1 2026 earnings call, SEC filings, and Wall Street commentary, we break down the company’s ambitious turnaround strategy and the financial realities behind it.

The discussion explores Sweetgreen’s struggle to transform itself from a beloved but historically unprofitable salad chain into a scalable, durable, tech-enabled food platform. We examine the company’s alarming 12.8% comparable sales decline, ongoing operating losses, and razor-thin restaurant margins — while also analyzing the operational fixes management is implementing through its “Sweet Growth Transformation Plan.”

The episode dives into:

• Sweetgreen’s operational overhaul known as “Project One Best Way”
• The nationwide launch of wraps and their role in driving incremental customer traffic
• The importance of Sweetgreen’s direct digital ecosystem and loyalty strategy
• The company’s growing labor cost pressures and predictive staffing algorithms
• The Infinite Kitchen automation system and how it could reshape restaurant economics
• Why Sweetgreen sold its robotics company Spice to Wonder Group for $186.4 million
• How that sale transformed massive fixed R&D expenses into scalable variable costs
• The founder-controlled voting structure and what it means for investors

Most importantly, the episode challenges listeners to think critically about the future of modern restaurant businesses. If food brands outsource delivery logistics, kitchen automation, and operational infrastructure to third parties, where does the true enterprise value actually reside?

Is Sweetgreen becoming the future of food… or evolving into a highly branded real estate and customer acquisition company powered by external platforms?

This episode breaks down the numbers, the strategy, and the risks behind one of the market’s most polarizing restaurant growth stories.

Summary:

Artificial intelligence is creating an enormous energy problem — and nuclear power may be the only realistic solution.

In this episode, we break down NuScale Power (ticker: SMR) and its small modular reactor technology to see whether the company is building the future of off-grid AI infrastructure or burning cash chasing a dream.

We cover:

• why hyperscale AI data centers are driving nuclear demand
• NuScale’s massive regulatory advantage
• how their factory-built reactors work
• why their reactors can operate with minimal water and no traditional evacuation zone
• the company’s billion-dollar cash position
• the huge risks hidden inside their financial filings
• and the high-stakes bet on future nuclear commercialization

At the center of it all is one massive investor question:

If AI truly needs unlimited reliable power… could small modular nuclear reactors become the next trillion-dollar infrastructure shift?

Summary:

Advanced Micro Devices is transforming from a traditional chipmaker into a full-scale AI infrastructure company — and the scale is becoming almost unimaginable.

In Q1 2026, AMD posted $10.3 billion in revenue, up 38% year-over-year, while data center revenue surged 57% to $5.8 billion. But the real headline is Meta’s massive AI partnership with AMD involving up to 6 gigawatts of compute infrastructure — enough power to rival a small city.

The episode explains why the future of AI isn’t just about GPUs anymore. As “Agentic AI” systems grow more complex, CPUs become critical for orchestrating massive AI workloads. That’s why AMD doubled its projected server CPU market opportunity to over $120 billion by 2030.

We also break down:

• AMD’s mega-deals with Meta Platforms and OpenAI
• The EPYC and Instinct AI strategy
• Risks from gaming declines, export restrictions, and reliance on Taiwan Semiconductor Manufacturing Company
• Why the next AI bottleneck may not be chips… but electricity itself

The big question: Is AMD building the backbone of the AI economy — or entering an arms race the global power grid can’t sustain?

💻 Summary:

This podcast breaks down whether IonQ is becoming a true commercial leader in quantum computing or remains a speculative science experiment burning massive amounts of cash.

IonQ posted explosive 2025 growth:

• Revenue jumped 202% to $130 million
• Q4 revenue surged 429%
• Most revenue now comes from commercial customers instead of government grants

The company is pursuing a bold “quantum railroad” strategy — not just building quantum computers, but also selling critical infrastructure and components to the entire industry, including competitors. Massive acquisitions, especially semiconductor manufacturer SkyWater, are aimed at controlling the U.S. quantum supply chain.

The episode explains IonQ’s trapped-ion technology, its high-accuracy barium qubits, and the race toward building a 10,000 logical qubit machine capable of solving problems classical computers cannot.

But despite the hype, the risks are enormous:

• The company is still losing hundreds of millions annually
• Scaling quantum systems may prove far harder than expected
• AI-powered classical computers may improve so fast that quantum advantage never arrives

The core investor question:

Is IonQ building the future of computing — or betting billions on a technological breakthrough that may never fully materialize?

🎧 Podcast Episode Summary

What if you had to rebuild a rocket engine… while it’s already in orbit?

That’s the challenge facing IREN, a former Bitcoin mining company now attempting one of the most ambitious pivots in modern infrastructure: transforming into a full-scale AI cloud powerhouse.

In this episode, we break down IREN’s evolution through an investor lens—digging into earnings calls, SEC filings, and the mechanics behind a $9.7 billion Microsoft AI deal. What emerges is a story not about software, but about something far more scarce: power, land, and data center infrastructure.

We explore how rising Bitcoin mining costs—driven by the halving and increasing global competition—pushed IREN to rethink its business model. Instead of relying on volatile crypto revenues, the company is now redirecting its massive energy capacity toward AI workloads that offer stable, contracted cash flows.

At the center of this transformation is a strategic framework built on three pillars:

Capacity, Customers, and Capital.

With 4.5 gigawatts of secured power, innovative financing backed by Wall Street, and prepayments from Microsoft, IREN is positioning itself as a critical player in the AI supply chain.

But the opportunity comes with real risks. We unpack:

• Heavy reliance on a single hyperscale customer
• The logistical challenge of deploying 140,000 GPUs
• The fear of rapid hardware obsolescence
• Regulatory and grid constraints

We also dive into the key strategic choice facing the company:

Should they act like a landlord (co-location)… or a luxury hotel operator (owning and monetizing AI compute directly)?

Finally, we zoom out to a bigger idea shaping the future of tech:

“Time to data center” may be the most valuable asset in the AI era.

Because while software can be built overnight, power infrastructure cannot.

👉 The question we leave you with:

As AI demand explodes, will the companies that control electricity and infrastructure ultimately hold more power than the tech giants themselves?

🧠 Summary:

After 100 episodes, I’m pulling back the curtain.

In this milestone video, I share the 10 hardest money lessons it took me decades to truly understand—lessons about building wealth, investing smarter, avoiding costly mistakes, and finding meaning beyond money.

From the brutal truth that making money is hard, to why having a unique skill edge beats chasing “get rich quick” schemes, to the painful mistake I made with Bitcoin… these insights could save you years of trial and error.

You’ll learn:

* Why most people fail early (and how to push through)

* How to build a rare, high-value skill stack

* The mindset shift that separates wealth builders from everyone else

* Why being wrong is actually your biggest advantage

* The real purpose of money (and what actually makes a rich life)

This is not theory. These are hard-earned lessons from experience.

👉 If you’re serious about building wealth and living intentionally, this one matters.

Summary:

This episode argues that the real AI investment story in 2026 is not flashy AGI hype, but the rise of AI agents: software systems that do not just answer prompts, but can plan, use tools, access live company data, and take actions on behalf of people and businesses.

Using Google Cloud’s AI agent trends report as the backbone, the episode explains that companies are moving from instruction-based computing to intent-based computing. Instead of employees manually clicking through software, writing code, or running queries, they can state the outcome they want and let agents handle the execution. That shift can dramatically improve productivity, creating “10x employees” who orchestrate systems of specialized agents rather than doing every task by hand.

The discussion highlights how this changes business economics. Companies using agents can operate with fewer people doing more strategic work, which can widen margins and separate winners from legacy competitors. Real-world examples, like Suzano’s natural-language-to-SQL agent for SAP, show how agents can slash friction and unlock major efficiency gains across large organizations.

The episode also explores the infrastructure making this possible: A2A protocols for agents to work across departments, MCP to connect language models to live enterprise data, and AP2 for tightly controlled autonomous purchasing. Together, these systems enable “digital assembly lines” where agents detect problems, coordinate responses, and even complete transactions with minimal human intervention.

On the customer side, the podcast argues that modern “agentic concierges” are replacing old scripted chatbots with grounded, proactive service systems that understand company policies and live operational data. That idea extends to security, logistics, and commerce.

The big investing takeaway is that the true moat is not the model itself, since foundational AI will become commoditized. The real advantage lies in a company’s proprietary data, its ability to ground agents in that data, and its management team’s ability to drive adoption across the workforce. The episode closes by arguing that investors should stop rewarding AI theater and instead look for companies building grounded agent workflows, retraining employees, and creating measurable operating leverage from automation.