NYSE, ICONIQ, NFDG/ AI Grant: From SaaS to Superintelligence How AI Is Rewriting Growth Economics

Market Valuations and Traction

• Private market valuations for AI companies are at "unprecedented" levels, with several firms reaching $100M+ ARR in under a year.
• Reported private valuations include OpenAI at $300 billion, Anthropic at $60 billion (expecting higher price in next raise), Cursor near $20 billion, and Perplexity at $14 billion.
• Most of these top-valued entities are only two to three years old, creating a valuation environment where growth metrics supersede traditional age-based benchmarks.
• Hirsch (NFDG) notes that investment logic focuses on finding "100x winners" rather than 2x certainty, justifying aggressive multiple discounts based on potential parabolic growth trajectories.

Valuation Metrics and Financial Mechanics

• Traditional KPIs (scale, growth rate, burn, retention) remain relevant, but the definition of "great" has shifted upward, rendering historical multiples uncomfortable.
• Unit Economics:
  • AI companies can sustain 80%+ gross margins similar to traditional SaaS; lower reported margins (10–20%) often stem from including free-tier compute in COGS rather than Sales & Marketing.
  • High inference costs (API calls) compress gross margins for AI-native apps where the product core relies heavily on external model usage.
  • Infrastructure providers and AI-native apps face heavy CapEx and R&D costs for training models, requiring a flywheel effect to amortize these expenses.
• Pricing Models:
  • Hybrid pricing (combining seat-based and usage-based) is emerging as the dominant model, with ~40% of AI builders surveyed using this approach.
  • Pure usage-based pricing is common, but "outcome-based" pricing is rare (~6% of surveyed companies).
  • Inference costs are decaying rapidly (~10x improvement annually) due to hardware cycles (Nvidia, AMD) and algorithmic efficiency, allowing companies to absorb high usage while improving unit economics over time.

Strategic Investment Focus: AI-Native vs. AI-Enabled

• Cloud-to-AI Transition: Iconic views the current shift as the next major technological transition following the on-prem to cloud era, favoring companies with modern architectures that can easily embed AI.
• Portfolio Examples: Iconic cites Intercom (via "Finn") as an example of a legacy SaaS company successfully pivoting to become an AI-native entity.
• Foundational vs. Application Layers:
  • Foundation model builders (e.g., OpenAI, Anthropic) focus on technical elegance and APIs, generally avoiding "schlep-heavy" vertical applications (e.g., customer service implementation).
  • Application-layer companies defend themselves through deep workflow integrations, trust, and accuracy rather than just the underlying model.
  • Infrastructure tools (e.g., vector databases) face the risk of being "eaten" by larger players building broader toolkits, necessitating clear "act two" strategies.

Defensibility and Competitive Moats

• Core Defense: Hirsch argues the only viable defense is "great offense" defined by velocity, shipping speed, and customer iteration; data advantages (via fine-tuning/RFT) are secondary.
• Vertical Specificity: Defensibility is strongest in verticals where hyperscalers lack core competency, such as Office of the CFO, legal, and healthcare.
• Market Dynamics: The market is described as highly competitive where "entropy takes over" without continuous innovation, but "steamrollers" (hyperscalers) cannot cover every narrow niche.

Forward-Looking Opportunities (3–5 Year Horizon)

• Biology and Health: Hirsch identifies AI/ML intersection with biology as a high-impact sector, citing GLP-1s as a precursor to AI-driven drug discovery and genomic analysis (DNA, RNA, proteomics) potentially shortening R&D cycles.
• Vertical Software Disruption: Iconic anticipates massive disruption in offline industries (healthcare, legal, finance, agriculture) moving to "Service as Software," where AI solves fundamental labor shortages and operational inefficiencies.
• Infrastructure Evolution: New categories are emerging daily, requiring builders to navigate a landscape where tools may be commoditized by broader platform offerings.

Speaker Profiles and Firms

• Hirsch (NFDG): Multi-stage investor (seed to growth) with an early-stage accelerator ("AI Grant") and proprietary GPU cloud ("Andromeda Cluster"). Active in AI since early 2022.
• Sarah (Iconic): Principal at a $21B venture and growth fund based in London, covering product-market fit through IPO. Focuses on early AI-native startups and cloud-native SaaS companies integrating AI.