The Speed of Thought

When execution stops being the bottleneck, the entire machinery of software development becomes dead weight. We wrote recently that The Backlog Is Dead, but the death of the backlog is just one symptom of a deeper structural shift. The distance between wanting a piece of software and having it is shrinking fast enough to change what the work feels like.

Imagine an AI that could build exactly what you ask for the moment you ask for it. No queue, no prioritization, no backlog. You think it, it exists. That is still a thought experiment, but it is useful because it gives us a direction to measure against. Now scale it back a little: the AI is not instant, but it is fast, and you can run many copies at once. Most tasks finish before you have mentally moved on from defining them. That is much closer to where we are right now.

We can already spin up dozens of agents in parallel to chew through different parts of the Ikigai codebase at the same time. They are not infinitely fast, and they still need our supervision, but they are fast enough to change the physics of the work. The friction has shifted from how long something takes to build to whether we can clearly articulate what we actually want.

Digital creation is approaching the speed of thought. Not literal telepathy. Just a world where execution is getting close enough to human intent, fast enough and cheap enough, that the old delays no longer dominate the process.

We covered the practical side of this shift already: estimation, sprint planning, specialization boundaries, all the rituals built on the assumption that execution is scarce. Once agents can cross those boundaries without breaking stride, the logic holding the rituals together dissolves. The usual objections are still valid. Models make mistakes, miss subtle context, and require iteration. We see those failures every day in our own logs. But those problems describe the current friction in the system, not a permanent speed limit. The gap is shrinking on a timeline measured in months.

The more interesting question is what replaces the old machinery. Planning does not disappear, but it changes shape. “What can we afford to do this quarter?” becomes “What is worth asking for?” When the cost of trying something collapses, judgment matters more than scheduling. Clarity of intent matters more than estimation accuracy. Taste matters more than throughput.

Building Ikigai has forced us to abandon practices that made perfect sense a year ago. The backlog did not die because we found a better planning ritual. It died because execution stopped being the scarce resource around which everything else had to be organized. The future does not arrive all at once. It shows up first as a handful of old habits that stop making sense. For us, the backlog was one of them.

Co-authored by Mike Greenly and Claude Code