According to @nic__carter, AI inference costs are falling 10–1000x annually, implying that a $200/month AI subscription could drop by roughly half within six months as VC subsidies bridge near-term pricing (source: @nic__carter on X, Aug 15, 2025). For traders, this points to price compression risk across AI SaaS and model API vendors, warranting conservative revenue and ARPU assumptions in the near term (source: @nic__carter). The broader downtrend is corroborated by vendor pricing, as OpenAI launched GPT-4o at $5 per 1M input tokens and $15 per 1M output tokens in May 2024, below prior GPT-4 Turbo levels, reinforcing rapid cost deflation in the stack (source: OpenAI, May 13, 2024). In crypto markets, lower inference costs could expand adoption of onchain AI agents and data feeds while pressuring compute-linked revenue models, prompting a valuation re-rate toward usage-driven growth over pure pricing power, based on Carter’s cost trajectory (source: @nic__carter). Near-term trade tilt: prioritize adoption and volume beneficiaries over high-price narratives until pricing stabilizes, given Carter’s projected deflation path (source: @nic__carter).