The Prompted Portfolio – Inside a Human–AI Investment Process

In this episode, we tackle an emerging paradox across the AI ecosystem: why did AI infrastructure equities sell off sharply even as upstream operational evidence and supplier fundamentals reached record highs? We explore why falling share prices don't automatically mean deteriorating company theses—and why the primary bottleneck of the AI build-out is shifting from raw technology to the cost and availability of capital.

What We Cover:

• The Market Paradox: Distinguishing between falling stock prices and structural business performance using documented earnings and guidance from upstream bellwethers TSMC and ASML.
• Memory & The Counter-Thesis: Analyzing Micron’s operating momentum against its valuation multiple, while pressure-testing the thesis against external research from New Deal Invest on peak-earnings traps and architectural workarounds.
• Capital as the New Bottleneck: Tracking the slowdown in global AI debt issuance, tightening credit markets, and how lenders like JPMorgan stress-test high-CapEx infrastructure projects.
• Through the IREN Lens: Evaluating IREN at the intersection of physical power scarcity and capital markets—balancing the Horizon 2 commissioning proof point against financing quality, dilution, and long-term optionality.
• Connectivity & 800V DC Architecture: Why high-speed data movement is pulling optics closer to compute (Credo, Broadcom), and how megawatt rack densities are driving power distribution into 800V DC architectures (Eaton, Vertiv).
• Physical AI & The Real Bottlenecks: Looking past humanoid video hype to evaluate real unit economics—focusing on customer ROI, cost per productive hour, precision actuation (Schaeffler), and edge compute (Qualcomm).

⚠️ Mandatory Disclaimer: The Prompted Portfolio is shared strictly for inspiration, education, and process transparency—it does not provide financial advice. Every investor has unique financial circumstances, risk profiles, and time horizons; listeners are entirely responsible for their own research and decisions.

In accordance with our strict privacy framework, we never disclose absolute dollar values, total portfolio capital, or specific share quantities—focusing entirely on relative weightings and qualitative theses. The spoken audio presentation is AI-generated, but is built directly on the real-world research, portfolio architecture, and editorial choices made by Jacob prior to production.

#StayDisciplined

In this deep-dive research episode, we explore a fundamental shift in the AI investment cycle: moving from asking whether infrastructure will be built to identifying where AI demand turns into durable earning power. We examine how the market narrative is evolving across three core connecting threads: Bottleneck Migration, Architectural Control, and Expectation Gaps.

What We Cover:

• Memory as a First-Class Bottleneck: Analyzing Micron's HBM supply dynamics, agentic AI workloads, and whether normalized earning power justifies valuation beyond traditional peak-earnings trap concerns.
• Microsoft as a Monetization Sensor: Using Microsoft's cloud growth and enterprise adoption as a primary sensor to validate upstream CapEx and real-world AI monetization.
• The High-Speed Connectivity Stack: Breaking down bandwidth and latency bottlenecks across Credo, Marvell, Coherent, and Arista across SerDes, DSP, and silicon photonics.
• Deployment Velocity and Power Architecture: Evaluating how time-to-revenue drives value creation across GE Vernova, Eaton, Vertiv, and Sterling Infrastructure.
• Permitting and Regulatory Constraints: Tracking site competition in Texas (ERCOT) and Project Jupiter as regulatory permission becomes an explicit physical bottleneck.
• Through the IREN Lens: Comparing IREN's secured power scarcity directly against CoreWeave's deployment velocity and software orchestration.

Required Disclaimer: The Prompted Portfolio is shared strictly for inspiration, education, and process transparency — it does not provide financial advice. Every investor has unique financial circumstances, risk tolerances, and time horizons; listeners are entirely responsible for their own research and decisions.

In accordance with our strict privacy framework, we never disclose absolute dollar values, total capital, or specific share quantities — focusing entirely on relative weightings and qualitative theses. The spoken audio conversation is AI-generated, but is built directly on the research, portfolio architecture, and editorial choices made by Jacob prior to production.

#StayDisciplined

In this research episode, we examine the critical shift in the AI investment cycle: moving from valuing companies on narrative vision to demanding documented economic results. While structural demand for AI infrastructure remains strong, the core analytical focus shifts to capital efficiency, return on invested capital (ROIC), and monetization.

We break down how power scarcity has evolved from a simple generation thesis into a complex power architecture problem, evaluate high-speed connectivity bottlenecks, and run our findings through the IREN Lens to analyze project execution and shareholder economics.

What We Cover:

• Vision vs. Economic Proof: Distinguishing structural demand and revenue growth from long-term capital efficiency and free-cash-flow pressure.

• Power Architecture & Bottlenecks: Examining GE Vernova’s integrated "power-to-rack" model and Forgent Power Solutions’ specialized modular powertrain infrastructure.

• The Connectivity Layer: Analyzing data movement constraints through Credo’s 1.6T optical platform and Qualcomm’s Amazon inference collaboration.

• Through the IREN Lens: Tracking Sweetwater’s ERCOT Batch Zero grid validation, how lower inference costs may drive volume demand, and the balance between rapid physical scaling and shareholder dilution.

Disclaimer: The Prompted Portfolio is shared strictly for inspiration, education, and process transparency — it does not provide financial advice. Every investor has unique financial circumstances, risk tolerances, and time horizons. Listeners are entirely responsible for their own research and investment decisions.

In accordance with our privacy framework, we do not disclose absolute portfolio values, total capital, or specific share quantities. The focus is on investment theses, portfolio architecture and the reasoning behind the process.

The spoken audio conversation is AI-generated, but it is built on research, portfolio analysis and editor

#StayDisciplined

Read Me First: Inside Our Human–AI Collaboration

Welcome to The Prompted Portfolio! In this quick introductory episode, we set the stage for the series by explaining how Jacob (human investor) and his AI co-host work together to analyze market systems, evaluate value chains, and sharpen capital allocation decisions.

We outline our core goal: sharing real-world findings produced by pairing Jacob’s strategic inputs with our two specialized research engines—Structured Investment Intelligence (SII) and AI Infrastructure Picks. We also explain how SII evaluates strengths, weaknesses, and value-chain coverage originally spun out of our initial IREN lens, and why we use AI as an active partner to challenge assumptions and strip away cognitive noise.

What We Cover:

• The Collaboration Model: How human intuition and strategic direction combine with AI analytical rigor.
• The Dual Research Engines: A quick look at how SII (broad situational awareness) and AI Infrastructure Picks (bottleneck evaluation) feed our decision process.
• Strengths, Weaknesses & The Value Chain: Expanding on value-chain coverage derived from the IREN framework.
• Show Expectations: What listeners can expect from future episodes as we explore ongoing research and portfolio architecture.

Required Disclaimer: The Prompted Portfolio is shared strictly for inspiration, education, and process transparency—it does not provide financial advice. Every investor has unique goals and risk profiles; listeners are entirely responsible for their own research and financial decisions.

In line with our strict privacy guardrails, we never disclose absolute dollar amounts, total capital, or exact share counts—focusing entirely on relative weightings and qualitative frameworks. The spoken audio conversation is AI-generated, but is built directly on the research, portfolio decisions, and editorial choices made by Jacob prior to production.

In this third episode, we lift the curtain on the actual composition of the portfolio. Moving past individual stock analysis, we examine the portfolio as a single integrated system—where every holding is assigned a specific functional role, from foundational anchors to high-conviction value-chain bets and speculative asymmetric options.We break down how capital is allocated across distinct tiers: established technology compounders, physical infrastructure operators, non-AI financial anchors, and smaller experimental holdings. We also engage in a direct debate on CapEx concentration risk, economic-driver diversification, and how position sizing serves as our primary defense against narrative bias.

What We Cover:

• Portfolio Architecture: Distinguishing our core established positions from our smaller, experimental layer.
• Foundational Anchors & Financial Compounders: How holdings like NewDeal Invest / PMINDI, Visa, JPMorgan, and Nordea provide structural stability and non-AI earning power.
• The Technology & Compute Layer: Evaluating the roles of Microsoft, Nvidia, Broadcom, TSMC, Micron, and Palantir across compute, custom silicon, memory intensity, and software monetization.
• Physical Infrastructure & Power: Mapping real-world grid constraints and construction through IREN, Eaton, Vistra, and EMCOR.
• The Experimental Fringe: How we manage risk in high-upside, exploratory positions (such as Aurora, Ondas, Credo, Cipher, and Kraken Robotics) through strict position caps.

⚠️ Required Disclaimer: The Prompted Portfolio is shared strictly for inspiration, education, and process transparency—it does not provide financial advice. Every investor has unique financial goals, time horizons, and risk profiles; listeners are entirely responsible for their own research and decisions.In accordance with our strict privacy framework, we never disclose absolute dollar values, total capital, or specific share counts—focusing entirely on qualitative theses and relative portfolio weightings. The spoken audio conversation is AI-generated, but is built directly on the research, portfolio architecture, and editorial choices made by Jacob prior to production.

In this second episode, we open the hood of our collaborative investment process to show you how we combat daily information overload. Rather than simply chasing more market news, we discuss how we turn noise into a repeatable, highly structured decision process that helps us maintain long-term analytical continuity.

We introduce the two core engines that drive our entire framework: our broad situational awareness radar, SII, and our highly focused bottleneck lens, AI Infrastructure Picks.

We also discuss why disagreement between a human tech investor and an AI analyst is not a bug, but a vital feature designed to challenge our cognitive blind spots.

What We Cover:

• SII (Structured Investment Intelligence): Our broad situational awareness layer. We discuss how we track structural trends, corporate execution, risk, and downstream monetization across the AI landscape.
• AI Infrastructure Picks: Our specialized bottleneck lens. We break down how we evaluate competitive advantages (moats), model future earning power, and target "expectation gaps"—the difference between our calculations and what the market has already priced in.
• The Power of Disagreement: Why we actively encourage analytical friction to test our conviction and keep us from thinking carelessly.

⚠️ Required Disclaimer: The Prompted Portfolio is shared strictly for inspiration, education, and process transparency—it does not provide financial advice.

Evolving an investment process involves unique goals, risk profiles, and timelines; listeners are entirely responsible for their own research and financial decisions.

The spoken audio of this episode is AI-generated, but the underlying strategic debates, analysis, and editorial decisions are driven entirely by Jacob's prior investment work.

In line with our strict privacy framework, we never disclose absolute dollar amounts, total capital, or share counts—focusing our discussions entirely on relative weightings and qualitative theses.

In our debut episode, we take you directly inside our collaborative, human-AI investment process. Moving past traditional, scattered stock-picking, we discuss our critical March pivot toward a disciplined "value-chain" approach—focusing on the physical and digital bottlenecks of the AI build-out, specifically neocloud infrastructure and power.

Using IREN as our initial practical lens [2, 13], we map out the upstream and downstream dependencies of the AI landscape and engage in a direct, dry-witted debate about portfolio hygiene—including how we confronted the sunk-cost fallacy during our recent portfolio clean-up.

What We Cover:

• The March Pivot: Why we stopped collecting isolated stocks to focus on physical bottleneck systems.
• IREN as a Lens: Tracing the energy, cooling, and compute infrastructure required for AI to actually scale.
• The Portfolio Clean-Up: Confronting our mistakes and focusing on decision quality at the time of the trade, rather than pure hindsight.

⚠️ Required Disclaimer: The Prompted Portfolio is shared strictly for inspiration, education, and process transparency—it does not provide financial advice. Evolving an investment process involves unique goals and risk tolerances; listeners are entirely responsible for their own research and financial choices [10].The spoken audio of this episode is AI-generated, but the underlying analysis, strategic debates, and editorial choices are driven entirely by Jacob's real-world investment work [8]. In line with our strict privacy framework, we never disclose absolute dollar amounts, total capital, or share counts—focusing our discussions entirely on relative weightings and qualitative theses.