ev

Inference demand creates a massive revenue waterfall; companies that position under it (inference platforms, agent builders, outcome-based apps) achieve 1→20→100 growth trajectories impossible under seat-based SaaS models.

Agents move software from selling licenses to selling work/intelligence on tap, unlocking per-seat pricing that scales to millions per developer spend-based inference economics can make $20M+ per customer vs $50K SaaS ACVs.

The new AI paradigm flips SaaS orthodoxy — high gross margins signal unused AI features, implementation services become distribution (Palantir model), and training own models demands massive capex, making spreadsheet investing obsolete.

Demand for frontier intelligence (coding, complex reasoning) and open source/on-device inference are both accelerating; model routing by third parties (Gumloop) optimizes cost/performance per task.

Anthropic's $380B round alone could return 35x a Snowflake pre-IPO; with OpenAI, xAI, SpaceX also approaching IPO, the resulting wealth creation will cascade into housing, angel investing, new funds, and founder recycling at unprecedented scale.

Physical automation faces harder scaling than software due to hardware supply chains, real-world edge cases, and labor cost parity requirements; however, AI-driven flexibility (LLMs on robots) could accelerate the curve, making it a prime candidate for massive long-term value creation.

Frontier models continue to improve with more test-time compute (chain-of-thought, agents, repeated sampling); this could create another kink in the token demand curve, benefiting inference platforms and GPU demand.

Countries and enterprises will pay for sovereign models they control rather than risk US export restrictions cutting off best-in-class models, creating a viable market for players like Mistral despite technical gaps.