Bitcoin's Next Move Depends On One Fed Decision | Jordi Visser

— AI mid-cycle slowdown is driving a deleveraging event across tech stocks; many semiconductor and AI names have retraced 30–60% from peaks after 3–10x gains, though Jordi expects consolidation rather than complete collapse.

— Open-source vs. frontier models: Kimi K3 performs competitively, but enterprise adoption will likely favor U.S. closed-source models (Anthropic, OpenAI) due to integration, compliance, and cultural alignment concerns; "cultural weights" in models remain poorly understood and will become critical.

— Model routers and multi-model inference are emerging as enterprises optimize costs by routing simple queries to cheaper models and complex ones to frontier systems; the optimal number of models in such systems remains unclear.

— Inflation has cooled significantly; PCE core data is an outlier. Fed Chair Kevin Warsh signals reform-focused, AI-aware monetary policy rather than traditional hawkish or dovish stances, reducing July rate-hike odds to ~10%.

— Ethereum outperforming Bitcoin (up ~20% month-to-date); crypto is attracting capital due to lower realized volatility (~30) versus AI stocks (~100), offering better risk-adjusted exposure on a vol-adjusted basis.

— AGI convergence and disruption timeline: If AGI arrives within three years, public companies face structural multiple compression despite strong earnings growth; AI-native private firms and robotics/embodied AI will likely outperform legacy public equities.

— Vol-adjusted positioning: Bitcoin's 30 vol versus AI's 100 vol means you can hold 3× as much Bitcoin per unit risk. Consider rotating or rebalancing toward crypto if your AI holdings have concentrated duration risk.

— Memory semiconductors and infrastructure plays remain attractive if earnings haven't yet "stepped up" materially; Micron, ASML, and similar names face longer runways than already-peaked players like NVIDIA, which now trades only 48% above June 2024 levels despite continued dominance.

— Avoid SaaS software long-term due to structural AI disruption; favor AI infrastructure (chips, compute, memory) and private/AI-native firms over legacy public software. Watch for robotics/embodied AI as the next major trade once consumer agents gain traction.