Possible After SaaS
70 snips
Mar 25, 2026 They debate whether traditional seat-based SaaS is collapsing or simply transforming under AI-driven customization. They explore token-based pricing, tuned AI systems, and how software will become more integrated and dynamic. They consider shifting engineering roles toward orchestration, verification, and monitoring. They unpack new sources of defensibility like network effects and token/compute economics.
AI Snips
Chapters
Transcript
Episode notes
Seat Based SaaS Model Is Becoming Unsustainable
- Traditional seat-based SaaS economics (large upfront engineering, stable recurring seat fees) are breaking because AI can generate customized systems per company.
- Reid Hoffman argues software sales won't vanish but must embed AI generativity, tuned models, and libraries to remain defensible.
Retrain Engineers To Orchestrate And Verify AI
- Engineers should shift from typing detailed code to directing AI: set goals, choose architectures, and verify outputs rather than handcraft every API call.
- Reid recommends mastering AI tooling, orchestration, monitoring, and fallback design as new core skills.
Greenfield Teams Gain The Biggest Immediate AI Boost
- New projects built from scratch see huge acceleration from current AI, while refactoring large legacy codebases remains slow and brittle.
- Reid notes AI struggles with massive existing code, so greenfield teams gain immediate productivity advantages.