First: Investment Decisions Are Your Own Responsibility
This article explains how to read trends, not stock recommendations or trading advice. For concrete investing, consult a qualified professional and use your own judgment.
"AI-Related" Is Not Monolithic
"AI-related companies" tends to be lumped together, but revenue structures differ entirely. Classifying by the previous chapter's 5-layer value chain organizes it.
- Hardware layer: strong demand but swayed by the capex cycle
- Cloud layer: a structure where AI demand is added on top of existing business
- Model layer: huge development cost; the monetization path is the issue
- App layer: high growth expectation but intense competition and shakeout
Reading Without Being Swayed by News
- Look at primary sources: earnings materials, official releases. Don't judge by social summaries or AI hearsay
- Separate expectation and results: stock prices price in future expectations. Announcement ≠ immediate rise
- Be cycle-aware: there are waves of capex, hype, disillusionment, settling
Organizing Information with AI
Summarize this earnings release's key points neutrally into "revenue growth," "AI-related mentions," "management guidance," "risk factors." Don't make investment judgments.
Don't Believe AI Predictions
Even if you have AI produce stock-price predictions, they won't hit. AI's role is up to "quickly organizing earnings/disclosure key points." Treat preparing decision material and the decision itself as separate.
Latest moves (June 2026): record capital inflows into AI-related stocks
Alphabet pulled off a record-setting equity sale of over $85B. The first tranche was planned at $40B but oversubscription pushed it to $45B (Berkshire Hathaway took $10B). Another ~$40B tranche is planned next quarter, taking the total to $85B — bigger than Petrobras’s 2010 record. CEO Sundar Pichai pegged 2026 capex at $180–190B, mostly for AI infrastructure and data centers. Cloud-layer hyperscalers continue to soak up the capital, which in turn supports the hardware layer (NVIDIA et al.).
Anthropic formally filed a confidential IPO with the SEC (Form S-1, June 1 PT). Multiple outlets call it a candidate for the largest IPO ever, coming just days after the May 29 Series H raise of ~$65B at a ~$1T valuation. The combination of Alphabet’s mega equity sale and Anthropic’s IPO preparation makes it clear that institutional investors are broadly ready to deploy capital into AI. The long-running question of whether "model layer" startups can access public-market capital is now opening up, starting with Anthropic.
Reading the news without being whipsawed (update): the two principles in this chapter — "separate expectations from results" and "be cycle-aware" — still apply. Alphabet’s equity sale signals strong investor demand for AI-related names, but stock prices already price in 1–2 years of forward expectations. Anthropic’s IPO is only at the filing stage; the listing date, offer price, and actual financials (gross margin, inference costs, churn) will only become visible in subsequent S-1 amendments — don’t form opinions based on headlines alone.
