Argues feet-on-the-ground research is essential to navigate the environment, understand demand patterns in Taiwan/Korea, and assess supply chain issues.

Acknowledges a 'white-knuckle environment' for tech but cites robust demand in Taiwan for Nvidia, memory, and components as bullish for hyperscalers.

States the nervous backdrop will pass and tech is where to be positioned, especially during sell-offs.

Calls software sentiment the most negative in over a decade, a 'massively disconnected narrative' from the AI use-case reality for companies like Microsoft, Oracle, Salesforce.

Believes these stocks are 'on massive sale' relative to their role in the AI revolution.

Says the 'AI goose trade' (fear of LLMs like Anthropic unseating software) is overblown for entrenched players like Salesforce, ServiceNow, Oracle, and cybersecurity firms.

Cites customer checks on architecture and use-case building as evidence.

States Microsoft's multiples on FCF and earnings are at multi-year lows. Margins will support mid-to-high teens FCF growth for 2-3 years.

Provides explicit price targets: $500-550 if 10% of base upgrades to AI (bear case), $600-650 in base case. Calls it 'the most disconnected stock in tech'.

Predicts cybersecurity budgets will rise from 5% to 10% of IT spend over 2-3 years due to expanded attack surface from AI agents.

Names CrowdStrike, Palo Alto, Checkpoint as beneficiaries, stating this opportunity is strategic and won't be unseated by AI models.

Reports demand is accelerating quarter-over-quarter, deals are getting larger, and there are memory/component shortages, signaling a 'memory super cycle'.

Frames this as 'year 3 of an 8 to 10 year buildout' for the AI-driven 'fourth industrial revolution', unchanged by geopolitical tensions.

Provides Apple price targets: $325 worst case, $375-400 best case. Believes Apple 'won by accident' by not spending capex internally and will monetize AI via Gemini, driving a $75-100 upside.

Calls the current sell-off excessive.