Top 10 Posts of 2025

Why It Matters

The divergence between massive infrastructure spend and plunging AI costs forces investors and founders to rethink financing structures, pricing models, and growth metrics, reshaping the future of the AI and SaaS markets.

Key Takeaways

• OpenAI commits $1.15 trillion compute contracts through 2035.
• DeepSeek reduces training costs over 90% with V3.
• Nvidia's $110B financing exceeds telecom bubble ratios.
• Secondaries supplied 71% of 2024 exit capital.
• Public SaaS growth fell to 17%, IPO drought persists.

Pulse Analysis

AI’s cost curve collapsed again in 2025, juxtaposing OpenAI’s trillion‑dollar compute commitment with DeepSeek’s ultra‑efficient models that slash training expenses by more than 90%. This divergence signals a market where capital intensity no longer guarantees competitive advantage; instead, algorithmic efficiency and explainability become the primary differentiators. Companies that can deliver high‑performing models at a fraction of the traditional spend are poised to capture enterprise contracts, especially as regulatory scrutiny around model transparency intensifies.

Financing dynamics also shifted dramatically. Nvidia’s $110 billion vendor‑financing program dwarfs the telecom bubble of the early 2000s, yet its top customers generated $451 billion in operating cash flow in 2024, suggesting a sustainable revenue base. The scaling wall myth proved false as Gemini‑3’s performance leap demonstrated that pre‑training improvements still yield exponential gains. Investors are therefore rewarding firms that combine deep pockets with demonstrable cash‑flow generation, redefining what constitutes a viable AI infrastructure play.

The liquidity landscape reflects a permanent move toward secondary markets, with 71% of 2024 exit dollars sourced from private‑sale transactions. This mirrors a broader shift where public SaaS growth slowed to 17% and IPO pipelines dried up, rendering traditional public‑market metrics less informative. As venture capital adopts private‑equity‑style exit strategies, founders must prioritize sustainable ARR growth and flexible pricing—maximization, penetration, or skimming—to thrive in an environment where public validation is scarce but private capital remains abundant.