Slow feedback loops and expensive computing forced optimization for machines over people
When computing power was scarce and costly, we optimized for machine performance over people performance. This shaped the planning-heavy methodologies of early eras, not because extensive upfront planning was inherently superior, but because slow feedback loops and expensive mistakes made iteration economically impractical.
Extremely costly computing performance led us to optimize machine performance over people performance. By the Agile era, this was reversed: Now knowledge and innovation, not bits and bytes, drove business performance.
From Wild West To Agile - Jim Highsmith
| Era | Feedback speed | Interface | What this enabled |
|---|---|---|---|
| Wild West | Hours (batch) | Punch cards → printout | Heavy upfront planning (iteration too slow) |
| Structured | Minutes | Green screen terminals | Still documentation-heavy, but faster validation |
| Roots | Seconds | GUI, mice, modems | Rapid prototyping becomes viable |
| Agile | Instant | Touch, cloud, CI/CD | Continuous iteration, learning loops |
The shift from hours to seconds of feedback didn't just make iteration possible, it made heavy upfront planning wasteful. Why spend weeks documenting requirements when you can build and test in days?
Technology constraints amplified the perceived need for certainty. That's also why The increasing in uncertainty drove each era of software development practices.
On the other hand, Agile practices only make sense when iteration is cheap