The AI Podcast What VC's Are Looking For in AI Startups Today
Mar 3, 2026
Investors are shifting toward AI that actually completes tasks and builds proprietary data moats. They avoid thin workflow layers, generic wrappers, and shallow product depth. Automation and execution beat surface-level chat interfaces. The trend favors vertical expertise, unique data, and defensible integrations over easily copied tools.
AI Snips
Chapters
Transcript
Episode notes
Investors Want AI That Actually Completes Work
- VCs now prize AI that completes tasks rather than just adds a chat UI.
- Aaron Holiday highlights vertical AI infrastructure and products that 'actually complete something' as preferred investments.
Avoid Thin Workflow Layers
- Avoid building thin workflow layers or generic horizontal tools that agents can replicate quickly.
- Thin UI automation and surface-level analytics are seen as weak because large models can reproduce core value fast.
Data Moats Trump Vertical Without Uniqueness
- Proprietary data moats are increasingly required for vertical AI to be compelling to investors.
- F-Prime's Abdul Abdirhan says vertical software without unique data is no longer very attractive.