The "Sasspocalypse" narrative is overblown. Incumbent SaaS companies are not being replaced but are instead leveraging AI to increase value and raise prices. The most significant near-term disruption from AI in enterprise software is the dramatic reduction in switching costs between providers.
Significant value will accrue to the application layer, which aggregates increasingly specialized foundation models for specific tasks. While traditional moats like network effects still matter, proprietary, live data has emerged as a more powerful and durable form of defensibility in the AI era.
San Francisco maintains a significant network effect advantage for ambitious builders due to its density of talent and tacit knowledge. Other ecosystems like Tel Aviv succeed by forcing an international-first mindset due to small domestic markets, a potential weakness for UK-based startups.
The Series A remains the optimal investment stage, balancing signal with price. VCs should be highly flexible on valuation for top-tier companies but remain firm on ownership targets to align with their support model. The nature of market "subsidy" has shifted from ad spend (2021) to compute credits, a healthier form of customer acquisition.
The biggest opportunities for startups lie in creating new, native AI categories (e.g., AI companionship) that large corporations are structurally too risk-averse to build. The future UI is not exclusively chat-based; browse-based interfaces will persist for users who want to "spend time," not just save it.
Keep pulling the thread on Anish Acharya.