Interview, Fireside Chat
How Anthropic’s $100M Anthology Fund Works | Menlo Ventures
- Future venture success is projected to rely primarily on founder quality, hiring capabilities, and deep operational involvement rather than pre-existing market predictions, with deal volume remaining below five per partner annually to facilitate M&A motions for underperforming assets.
- A new $100 million fund will be deployed with Anthropic to support seed teams and strategic partners outside of a traditional corporate venture structure, while future investment decisions will prioritize solving specific hiring bottlenecks and infrastructure needs over broad AGI timeline predictions.
- Economic models are expected to restructure over the next 10 to 50 years due to global population decline, necessitating faster product cycles and increased per capita consumption to sustain activity despite anticipated job displacement and labor shifts toward reduced non-productive headcounts.
- Technology demand will grow to address shortages in blue-collar and non-Western desk jobs, while significant information asymmetries between the AI community and practical sectors like insurance, trucking, and logistics are expected to persist, creating opportunities for bridging frontier research gaps outside academic funding.
- The AI investment landscape will face a persistent "30 premium" for Series A "AI-enabled" companies due to a constrained talent pool, alongside a distinction between genuine deep technical stacks and companies merely adjusting LLM prompts, with capabilities evolving from internet data training to reinforcement learning and test-time compute paradigms.
- Enterprise products failing to adopt product-led growth dynamics risk elimination by competitors leveraging viral mechanisms, while founders securing excessive capital may suffer reduced velocity, employee hunger, and recruitment difficulties due to limited upside perceptions at high valuations.
- Seed investments at billion-dollar valuations are generally deemed unjustifiable unless structured as SPVs with minimal risk, and long-term returns will increasingly depend on identifying founders who can hire in saturated markets and focus on singular product validation questions rather than early defensibility analysis.
- Progress in AI will be measured by an "economic tiering test" determining value generation relative to human labor, and while new technologies may cause temporary displacement, historical trends suggest humanity will adapt to new employment areas over the long term.