Browse Comments — Clean (de-noised)
Close reading of the corpus at each pipeline stage: raw → clean → relevant → coded.
81
comments matched
· page 3 of 5
Good prompting matters, but clear thinking and system design matter even more over time.
The craftsmanship analogy fits because strong software is usually built through repeated refinement instead of one perfect pass.
Crikey.... I would like to see AI take over that production...
Good reminder that craftsmanship still matters in modern coding workflows.
Loved the line about being the architect and reviewer. So true.
Claude Code is amazing, but direction and clarity still come from humans.
Very good framing of AI coding
Nice one 😀
Maybe #fakes of the best soccer balls😉 #SafetySandals are a minimum requirement
The soccer ball analogy is honestly perfect. Great products are built piece by piece.
Claude Code isn’t magic; it’s a workflow.
Iain Eyre Indeed true! Realizing that with my current project.
Hell of a process but it’s so worth it in the end ❤️
I will say this. I can understand this analogy - for sure. It makes sense (as do other similar analogies about AI and process and experience and what not). However, I am also wondering if maybe we shouldn't gloss over that, at least with this example, maybe the "process" is also underplaying an imbalance in the means of production? I am not trying to sidetrack this analogy too much (and I can see a lot of AI responses are happy to just nod and agree) - but I just thought that if we are going to look at AI as a process (which it very much is) - maybe we should also account for the idea that this process should also account for the equity of those (and the reality for those) who are and will be directly impacted by this technology. We have a very real opportunity to create "the good" with AI - and for sure, it can be a real level setting technology. But as is often the case, sometimes these revolutions come at the expense (time, effort, etc) of others. To have an understanding without the realization of the humanity behind it can be irresponsible (at the very least) and hurtful at expense of progress on the other side.
This explains the process really well because AI works best as a collaborator, not a shortcut. Strong systems are built through testing, refining, and thoughtful decision making.
the craftsmanship frame holds for maybe 6 more months. what feels like skill right now is mostly attention paid to a non-deterministic system. next model gen collapses 80 percent of that 'craft' into a single prompt, same way prompt engineering already collapsed.
People often underestimate how much structure matters when working with AI tools. When you treat it like a craft instead of a shortcut, the quality changes completely.
You means the best FOOTBALLS, that’s the sports name not soccer 🤣
The vending machine mindset is exactly why most people bounce off these tools after a week.
Clear inputs and tight review loops make AI coding actually useful.