AI 'wrappers' evolve into full harnesses with memory and workflow rules
Early AI products started as simple wrappers around models, but over time they added industry tools, workflow rules, and memory — and in doing so became full harnesses. The legal space is a clear example: products like Harvey and Spellbook didn't just surface AI, they built specialized infrastructure around it for a specific professional context. Greg Isenberg sees this pattern repeating across many industries, with significant opportunity for founders who go deep into a vertical rather than staying at the wrapper layer.
Go deeper with Greg Isenberg
Ask anything about this, and Greg Isenberg answers from everything they've shared.
