
"Software is what's left over after a group of humans finishes negotiating with each other about what the system should do. The code matters, but it is the residue of the harder work, not the work itself. For fifty years the residue was expensive enough to keep our attention on it. With coding agents the cost has fallen far enough that we can see what's underneath: people trying to agree."
An experiment tested structured-generation algorithms against open-source counterparts by evaluating whether outputs match the right token distribution rather than merely accepting input strings. A short explanation to Codex led to a working first version, showing how quickly coding agents can produce functional code. Coding agents are already changing how individuals write, but skepticism remains about claims that individual productivity automatically makes the industry move much faster. Software quality and progress depend on collaboration, since software is the residue after groups negotiate what the system should do. For decades, tools reduced the cost of producing and reviewing code, but agent-driven automation lowers that residue cost enough to reveal the underlying challenge: people negotiating, agreeing, and communicating a shared understanding of what they are building.
Read at The typical set
Unable to calculate read time
Collection
[
|
...
]