ERA
EARLY COMPUTING
Learning the first hard truth of software: machines are literal. They reward clarity, expose assumptions, and make vague thinking expensive.
JASON//SYSTEM / STORY
THE MACHINES CHANGED.
THE INSTINCT DIDN'T.
Technologies come and go. The recurring work is more durable: understand what the machine can do, understand what people actually need, and connect the two with a system that survives reality.
THIS IS A COMPUTING STORY, NOT AN EMPLOYMENT TIMELINE.
Logic becomes behavior.
ERA
DESKTOP SOFTWARE
Applications feel complete and self-contained. The challenge is making a machine useful through interfaces, state and predictable behavior.
ERA
THE WEB ARRIVES
Software escapes the desktop. Distribution becomes immediate, information becomes linked, and the browser starts becoming an application platform.
ERA
WEB APPLICATIONS
Pages become products. Client state, servers, databases, APIs and user experience begin behaving like one continuous system.
ERA
CONNECTED SYSTEMS
The interesting work moves between applications: authentication, APIs, integrations, data flows and the places where one service depends on another.
ERA
CLOUD
Infrastructure becomes programmable. Systems gain elasticity, managed services and distributed failure modes that demand stronger operational thinking.
ERA
SAAS
Software is no longer finished when it ships. It is continuously operated, measured, supported, billed, secured and improved.
ERA
AUTOMATION
Software stops waiting for every click. Queues, jobs, workflows and rules begin carrying the repeated work on their own.
ERA
GENERATIVE AI
Natural language becomes a serious computing interface. The cost of producing code and content changes, but judgment, architecture and verification become more important.
ERA
AGENTS
Software begins selecting tools, carrying context and coordinating multi-step work. The core engineering questions shift toward state, authority, observability and trust.
ERA
NEXT
After decades of software evolution, the interesting question is no longer only what software can do. It is what happens when software can reason, choose tools, coordinate work and act.
CURRENTLY COMPILING...THE CONSTANT
The tools keep changing. The instinct is still to look at a difficult system and ask: how should this actually work?See the systems