[Latent Space LIVE @ NeurIPS] State of AI Startups 2025 — with Sarah Catanzaro, Amplify Partners

The merger wasn't a sign of weakness but a strategic consolidation to meet new, much higher revenue expectations for a successful IPO, which now exceed $600 million. Both companies were individually healthy and beating targets, merging to accelerate their path to liquidity in a demanding market.

The data catalog market struggled because major platforms like Snowflake and dbt absorbed discovery features, making standalone tools redundant for human users. The bigger, missed opportunity was creating catalogs for machine-to-machine interaction (e.g., microservices, agents) and solving for governance rather than just discoverability.

As AI's novelty fades, apps face high churn. The solution is personalization through memory and continual learning. This is a difficult systems problem because it requires a paradigm shift from today's stateless inference to a stateful model where weights are updated dynamically based on user interaction.

The best application-focused AI companies are born from a need to solve a hard research problem to deliver a superior user experience. This "application-pull" approach, seen in companies like Harvey (RAG) and Runway (models), creates a stronger moat than pursuing research for its own sake.

The current AI funding climate is characterized by massive seed rounds raised on long-term vision alone, with no concrete near-term plan. The process has become highly transactional, forcing investors to make decisions in under a week, preventing deep diligence or the formation of a true partnership with founders.

The trend of buying expensive, simulated Reinforcement Learning (RL) environments is misguided. The most effective and valuable training ground is the live application itself. Companies can achieve better results by using logs and traces from actual users, which provides the most accurate data for agent improvement.