Interview, Fireside Chat
Atlassian CEO, Mike Cannon-Brookes on Why Everything is Overvalued & Are We in an AI Bubble
- The current era is predicted to last three to five years, characterized by complete up and down turmoil, before the speaker anticipates looking back at this period as "crazy" ten years from now.
- The company expects to employ more engineers and software developers in five years than today, with a specific plan to hire more software engineering graduates this year than in the previous year, aiming to create far more, efficient software and far better technology through iterative development.
- The market is expected to experience a "verdant creative kind of Cambrian explosion" where many current products are vastly overvalued while others remain undervalued, with the value of the undervalued likely outweighing the overvalued, though the long-term monetary and fundamental value of AI business models remains unknown.
- A shift in pricing models is anticipated, moving away from pure consumption-based approaches toward a blended model that retains per-seat pricing while incorporating more consumption-based elements, with the industry eventually getting used to a mix of both.
- The speaker plans to form opinions on the technology landscape's evolution over the next three years while committing to reassess these choices approximately every quarter.
- AI business models are expected to improve in code manipulation, long-form code handling, and ingesting large code bases, with switching costs currently low but expected to build over time based on value delivered, data, workflows, familiarity, and interface.
- Margins in AI businesses are believed to be built on customer value delivered over time rather than immediate capture, with some current applications expected to disappear while others profit from fundamental changes.
- Interface design is predicted to evolve toward a combination of modern existing UI and prompt-based layers, as users will seek stability rather than dynamically changing interfaces, with good design becoming a more valuable scarce resource that underwrites switching costs.
- The company expects that customers will create far more apps and more apps per customer, with "10x engineers" potentially becoming "100x engineers," and graduates bringing new ideas that shake up the existing talent world.
- Atlassian must prove its position in the AI era over the next five to ten years and needs to survive technological disruptions to become a multi-decade technology company, with a long-term goal of remaining a vibrant, competitive player by 2035.
- The speaker expects the speed of AI understanding and deployment to take longer than anticipated, while noting that founder-driven businesses must make real decisions and actions during this era of fast growth and low initial switching costs.
- New AI models are expected to arrive approximately every three months, with four to six companies competing in the foundational model space, while the industry may see a trend of businesses collapsing into agent-era interactions or adapting to "vibe coding" for non-technical users.
- The speaker believes that core technologists will build more platforms, that software creation is not output-bound, and that the high demand for software will drive the creation of significantly more and far better technology.