Productivity gains have paradoxically made software development harder by increasing expectations and project scale.
The Super Mario Bros. franchise serves as a case study, showing team sizes exploding from seven to over 400 people over decades.
The rebound effect and market competition compel companies to add more features, leading to ever-larger software applications.
Markets evolve from a 'pioneer phase' of experimentation to a structured 'market phase,' which in turn creates tools that enable new pioneers.
Continuous improvement sets ever-higher performance baselines, creating immense pressure and potentially unrealistic expectations for developers.
Technical solutions like asynchronous code or blockchain can introduce new, unintended complexities such as race conditions or market distrust.
Architectural patterns like microservices don't eliminate complexity but strategically move it to reduce cognitive load on individual developers.
As software scales, team roles become more specialized and hierarchical, mirroring architectural components as described by Conway's Law.
Relying on frameworks saves development time but can lead to vendor lock-in, outsourced knowledge, and deep layers of abstraction.
Ultimately, practical limitations like time and money prevent complexity from growing infinitely, forcing teams to find simpler solutions.
Jobs that call for the skills explored in this talk.