The Macro Shift: Exponential AI scaling laws are colliding with the slow, complex realities of institutional adaptation and capital cycles. The future of AI will be decided by this interaction, not just technical progress.
The Tactical Edge: Prioritize building solutions that abstract away institutional friction or offer clear, measurable value within existing, slower-moving frameworks. Focus on integration and governance, not just raw capability.
The Bottom Line: The next 6-12 months will test whether institutional inertia can be overcome by AI's capabilities or if architectural limitations around persistent learning will force a re-evaluation of current scaling assumptions.
The Macro Shift: Exponential technologies are driving a fundamental shift from scarcity-based systems to abundance, challenging the very definition of wealth and economic growth. This transition will be messy, marked by institutional resistance, but ultimately unstoppable.
The Tactical Edge: Cultivate a curiosity and exponential mindset, focusing on technologies with doubling patterns (AI, solar, biotech) and building solutions at near-zero cost. Position yourself to capitalize on the disruption of regulated, inefficient sectors.
The Bottom Line: The next decade will redefine societal structures and personal purpose. Embrace discomfort, learn relentlessly, and recognize that a future of radical abundance is not distant, but arriving in months, not years.
Evolution isn't solely random mutation; symbiogenesis, the fusion of cooperative entities, is a fundamental, overlooked engine of complexity and intelligence.
Design AI systems and decentralized networks with explicit mechanisms for "symbiogenesis" – allowing modules or agents to cooperatively fuse, forming higher-order, self-improving structures.
Recognizing life and intelligence as embodied computation, driven by fusion, offers a powerful new framework for building open-ended AI and understanding forces that drive complexity.
The Macro Shift: Insatiable AI demand meets the technical reality of rapidly depreciating model assets, pushing AI companies to prioritize infrastructure control and long-term capability scaling over short-term consumer-facing profitability.
The Tactical Edge: Invest in AI infrastructure plays (GPUs, energy, data centers) and companies building model-agnostic agentic systems, as these components offer more durable value than individual frontier models.
The Bottom Line: The market is underestimating the demand growth for increasingly capable AI models. Expect continued massive capex in compute, and position for a future where AI agents become indispensable, driving significant, sustained enterprise spend over the next 6-12 months.
AI's economic viability is shifting from model-specific gross margins to the long-term utility of persistent agents and the underlying compute infrastructure.
Invest in or build infrastructure plays (GPUs, energy, data centers) that support the insatiable demand for AI compute, recognizing that model software is a rapidly depreciating asset.
The market's recent "whiplash" on AI valuations misses the true demand growth and the strategic pivot towards infrastructure and long-running agents.
The AI industry is moving from a pure software-as-a-service model to a vertically integrated infrastructure play, where control over compute and power becomes the ultimate competitive advantage.
Invest in or build solutions that abstract away the underlying model, allowing for easy swapping between providers, while focusing on persistent agent memory and identity.
The market underestimates AI demand. Companies controlling infrastructure and delivering agents capable of sustained, high-value work will capture significant value over the next 6-12 months, even as model development costs remain high.
The AI industry is shifting from a pure software-like model to one where infrastructure ownership and continuous R&D are paramount.
Prioritize infrastructure investment: Given the GPU and energy constraints, securing or building proprietary compute infrastructure will be a decisive competitive advantage.
The next 6-12 months will see a continued capital expenditure arms race in AI infrastructure.
The AI industry is shifting from a software-like business model to one resembling capital-intensive infrastructure, where models are rapidly depreciating assets. This forces a focus on massive, continuous R&D and infrastructure buildout (GPUs, energy) to unlock future capabilities and markets, rather than immediate software-like margins.
Prioritize infrastructure investments. For builders, design systems with model agnosticism, allowing for easy swapping as models improve or become obsolete. For investors, evaluate AI companies not just on current gross margins, but on their ability to secure compute, attract top talent for R&D, and demonstrate a credible path to future market expansion through scale.
The next 6-12 months will see continued massive capital expenditure in AI infrastructure. Companies that can secure GPU supply and energy, while effectively managing the short lifespan of frontier models through continuous R&D, will hold a decisive competitive advantage. The market will increasingly reward long-term vision and infrastructure plays over short-term profitability.
If you look at how much they spent in R&D in the four months before they released GPT5, that quantity was likely larger than what they made in gross profits during the whole tenure of GPT5 and GPT5.2.
The models as a rapidly depreciating asset actually brings a little bit into focus of what might be the enduring asset... it seems to me that this part is infrastructure.
The market is always right... However, with that said, they didn't get the demand growth. They didn't get the way in which that demand is outstripping supply. They didn't get how much more we were going to demand as these models get better.
