Mike Cannon-Brookes expresses significant skepticism about the current AI market, describing it as a period of 'crazy' hype and 'vastly overvalued' companies. He critiques the circular flow of capital where VC funds flow to model companies, then to cloud providers, and back to the VCs' LPs, suggesting an unsustainable bubble.
Atlassian is deliberately avoiding building its own foundational model, instead opting to be a neutral platform that integrates and supports multiple competing models. This allows the company to leverage the best technology for specific use cases without being locked into a single ecosystem, focusing its resources on the application layer where it delivers direct value to customers.
Contrary to fears of AI replacing developers, Cannon-Brookes argues that AI tools will make engineers more productive, leading companies like Atlassian to hire even more of them to tackle a larger set of problems. He also foresees the rise of 'vibe coding,' where non-technical professionals can create simple applications, expanding the creator base without diminishing the need for specialized engineers.
Cannon-Brookes believes the underlying AI technology, particularly for tasks like code generation, will become a commodity with low switching costs. Consequently, sustainable competitive advantage will shift to other areas. He highlights good design as a key differentiator that will become more valuable and scarce as the technical barrier to creating software is lowered.
The traditional per-seat SaaS pricing model is likely to evolve in the AI era. Cannon-Brookes anticipates a shift towards blended models that incorporate metric or consumption-based pricing to account for the variable costs and value delivery of AI features, acknowledging that the final margin profiles for AI businesses are still uncertain.
Keep pulling the thread on Mike Cannon-Brookes.