We are all at zero now

I’m broken. I can’t think, I can’t develop. The development world has changed again, and I no longer know how it works. I do know that it still works, and that there is a new approach to software engineering emerging. I don’t know if, like prompt engineering, this is a temporary phase; I know it is not a fad. I also know that there definitely is something to it.

I brain dumped many thoughts on this, and playfully titled my thoughts “We are all at zero now”. But that’s not quite true. Some people have out right rejected the change that has come; some aren’t using AI at all yet. Not everyone has even gotten to zero. That is of course a good thing, but its also something important to note.

The zone is flooded with amateurs. And also with startups, who will soon die. And also with established companies, who may soon die. People are racing to ship, to find the seams and work within them; to find the new goals, apps, services, and ways of working. Platforms like Twitter are filled with engagement farming around The Way.

CLIs were a necessity, then a pleasure; many wanted to steer them towards humans, now others back towards LLMs. (LLM=true, 2>&1). With all the other approaches, I am unsure if this one is even worth it; the harness around claude, gemini, codex, will surely also chase this; should we just wait?

I think waiting is correct, but also wrong. Many but not all of us have anxiety about the world, but in this case I think it is seeing careers turned upside down, while also a seemingly growing movement that cares mostly to concentrate and protect wealth up. I lean strongly capitalist and free market, but I don’t see it trickling back down in its current form, even if I think the tech giants have the most to fear from these new advancements.

In a way this is good. Many have been going through this for a while; I believe part of the cultural split in Dems and Republicans in the US is the Dems failure to collectively grapple with this change, even as they support measures for leveling it out. The more people that are lumped into it, the harder it is for either party to ignore. Perhaps there is something good that will ultimately come of it. People need physical and economic security above all else; they will continue to throw the baby out with the bathwater in search of it.

But along the way, I don’t think we wait. I think we explore, hack, build. Its likely all for nothing, but there is still meaningful progress to make, knowledge to share, and an engineering culture to evolve. Life goes on whether in a form we like or not; we must all find our ways of fitting into it. I’m building more than ever, tinkering more than ever. Its harder than ever to finish an app, or a plan, or even a thought. This accentuates the anxiety, though I take solace in knowing that it is collective.

When you can’t see the end, its hard to know where to begin, where to focus your energy. For so long knowing the possibilities was easy, and from it making bets on where to invest your learning and your building was relatively easy. At worse you could scratch a personal itch, but at best turn your ideas into your career. But now? Everything is upside down. Agentic coding is great, but its hardly enough. Spec driven development, context engineering, harness engineering. This is where its moving, quickly. The players are integrating the concepts fast. The attempts are certainly amateur at time, but then seeing how simply connecting those attempts to reality - whether logs, metrics, data - and building playgrounds to experiment and get feedback… its easy to imagine how loops produce new forms of automation and changes.

At any rate, I guess in 2026 the slop fest is on. I hope we can clean it all up.