Why OpenAI and Anthropic Won't Win the App Layer | Glean Founder
20VC with Harry Stebbings · 1:00:08 · 6 days ago
Enterprise AI adoption is entering a phase where cost control and institutional ownership matter more than raw capability. While frontier models are powerful, businesses are increasingly shifting toward open-source options for financial sustainability, and founders should focus on building measurable value rather than attempting to replace human roles entirely.
- Frontier model dependence — Enterprises fear losing control over their institutional knowledge and operational data to external AI providers, necessitating a balance between convenience and autonomy .
- Open source shift — Roughly 90% of current business use cases can be satisfied by diverse, cheaper models, causing a move away from relying exclusively on expensive, closed-source providers .
- Operational costs — Many companies hit their annual AI budgets within months, pushing them to seek cheaper alternatives to maintain financial sustainability .
- ROI reality — Coding speed has surged, yet companies report that overall product shipping speed remains largely unchanged, making the return on investment difficult to quantify .
- The "replace yourself" trap — Encouraging staff to replace themselves with AI is a flawed objective, as current technology lacks the human judgment required to handle critical, nuanced decisions .
- Startup capital — The current overabundance of funding allows startups to make unsustainable financial decisions, such as paying inflated engineering salaries .
- Internal growth — Unlike peers who are shrinking headcount, the company plans to continue hiring to scale, viewing employees as essential assets for competitive advantage .