The AI industry is transitioning from a model-centric competition to an infrastructure and agent-centric one, where raw compute and persistent user experience dictate long-term value.
Prioritize investments in AI infrastructure providers and platforms that enable model agnosticism and agent memory.
Expect continued massive capital expenditure in AI infrastructure, a focus on enterprise solutions, and the rise of "sticky" AI agents that abstract away underlying model changes, shifting the competitive battleground.
The AI industry is moving from a software-like model, where products have long lifespans, to one where models are rapidly depreciating assets requiring continuous, heavy R&D investment.
Prioritize investments in AI infrastructure and agent orchestration layers that abstract away underlying models.
The market is underestimating the demand growth for increasingly capable AI models.
The Macro Shift: AI models are rapidly depreciating software assets, making the underlying compute and energy infrastructure the enduring value proposition.
The Tactical Edge: Prioritize building model-agnostic agentic workflows that retain memory and context, allowing for flexible model swapping and cost optimization.
The Bottom Line: The AI race is a capital-intensive marathon where infrastructure ownership and a long-term vision for capability expansion, not immediate model profitability, will determine market leadership over the next 6-12 months.
Invest in companies building core AI infrastructure (GPUs, energy, data centers) or those developing enterprise-grade AI agents that deliver measurable, long-duration value, rather than consumer-focused models with short lifespans.
The AI industry is moving from a software-like gross margin business to an infrastructure-heavy, capital-intensive play where sustained R&D investment is a prerequisite for market relevance, not just growth.
The market's recent jitters about AI capex miss the point: demand for increasingly capable AI is outstripping supply.
Prioritize investments in AI infrastructure plays (GPUs, energy, data centers) and companies building model-agnostic agent layers.
The market is underestimating the insatiable demand for increasingly capable AI, which will drive massive compute spend and make infrastructure the true bottleneck and value driver over the next 6-12 months.
Insatiable demand for ever-improving AI capabilities is driving unprecedented compute spend, but the true long-term value shifts from rapidly depreciating models to the underlying, enduring infrastructure and the persistent "memory" of AI agents.
Invest in or build solutions that abstract away the underlying model, focusing on agentic memory and robust infrastructure. This future-proofs against model obsolescence and capitalizes on the growing demand for persistent AI workers.
The market's recent "whiplash" on AI valuations misses the core truth: demand for advanced AI is outstripping supply. Companies that can build or secure infrastructure and develop sticky, agent-based experiences will capture significant value over the next 6-12 months, despite current profitability questions.
The AI industry is reorienting from a model-centric race to an infrastructure and agent-centric value proposition, where delivering persistent, high-value AI workers will outweigh the transient superiority of any single model.
Invest in or build solutions that abstract away the underlying LLM, focusing on agentic memory, workflow integration, and robust infrastructure.
The next 6-12 months will see a continued re-evaluation of AI valuations, favoring companies that demonstrate a clear path to monetizing agentic capabilities and owning critical compute infrastructure, rather than just shipping the "next best model."
The memory aspect of semiconductors today has gotten so extreme. Stuff is so expensive that people are simply not able to make lower-end equipment or like devices anymore. And this is like killing everything, right?
AI chips deliver 65% operating margins, exceeding gaming GPUs' 40%. This incentivizes NVIDIA to prioritize AI data center chips.
Meta's AI investments directly improve its core advertising business, generating substantial revenue from 3.5 billion users. This makes AI capex a straightforward investment.
The US is pivoting from a QE-fueled, government-led economy to a "free market" model under the new Fed Chair, Kevin Warsh. This means a potential reduction in the Fed's balance sheet (QT) and lower rates without yield curve control (YCC), leading to decreased US dollar liquidity.
Adopt a phased, data-driven allocation strategy. Michael Nato recommends an 80% cash position, deploying first into Bitcoin (65% target) at macro lows (around 65K-58K BTC, MVRV < 1, 200WMA touch), then into high-conviction core assets (20%), long-term holds (10%), and finally "hot sauce" (5%) during wealth creation.
The current "wealth destruction" phase, while painful, presents a rare opportunity to accumulate assets at generational lows, provided one understands the macro shifts and adheres to a disciplined, multi-stage deployment plan.
The financial world is splitting into two parallel systems: opaque TradFi and transparent onchain finance. Value is migrating to platforms that can simplify and distribute onchain financial products globally.
Invest in or build applications that prioritize mobile-native experiences, abstract away crypto complexities (like gas fees), and offer tangible real-world utility for onchain assets.
The future of finance is onchain, and "super apps" like Jupiter are building the necessary infrastructure and user experiences to onboard the next billion users.
Crypto's initial broad vision has narrowed to specific financial use cases, while AI and traditional markets capture broader attention. This means builders must focus on tangible value and investors on proven models.
Identify projects with novel token distribution models (like Cap's stablecoin airdrop) or those building consumer-friendly applications within new ecosystems (like Mega ETH) that address past tokenomics failures.
The industry is past its naive, speculative phase. Success hinges on practical applications, robust tokenomics, and competing with traditional finance, not just abstract ideals.
The Macro Shift: From unbridled, community-driven idealism to a pragmatic, business-focused approach. Early crypto imagined a world where "everything is a thing on Ethereum," but reality has narrowed its primary use cases to finance and trading, forcing a re-evaluation of tokenomics and community models. This shift is also driven by AI capturing mindshare and traditional finance co-opting blockchain tech.
The Tactical Edge: Re-evaluate token distribution models. Instead of relying on inflationary yield farming that creates sell pressure, explore innovative approaches like Cap's "stable drop" (airdropping stablecoins, then inviting participation in a token sale) to align incentives and attract long-term holders. Focus on building real products with defensible business models, even if they lean more "business" than "protocol."
The shift from centralized, static data aggregation to decentralized, real-time, incentivized intelligence networks is fundamentally changing how data-intensive industries operate.
Investigate subnet opportunities where incumbent data quality is low and validation is a core challenge.
The future of sales is not just about more leads, but smarter, fresher, and more relevant ones.
The Macro Shift: As trust erodes in traditional financial systems and geopolitical risks rise, capital is flowing towards more efficient, permissionless DeFi markets. This is forcing traditional finance to adapt or lose market share.
The Tactical Edge: Evaluate DATs trading below NAV for potential M&A or activist plays, as these discounts often reflect management misalignment rather than fundamental asset weakness.
The Bottom Line: The current market volatility, Fed policy shifts, and the rise of DeFi are not just noise; they are reshaping capital allocation. Investors and builders must understand these structural changes to position for the next cycle of institutional adoption.
