Hi~
I'm Mr.Ms.Dr.Prof.Sr.Bear.MMrr.. Best!

I've tried most coding agents, and they tend to share the same problem: during implementation, once an issue comes up, such as a function failing to do what it was meant to, the approach gets replaced, or an implementation detail just needs adjusting, the logic and data definitions added for the old approach rarely get withdrawn on their own. The agent simply starts a new implementation, writes another function, creates another class, and so on. Old and new approaches pile up until the code becomes a mess, and a human still has to go back and work out which functions, logic, and fields are no longer needed. An agent responsible for implementation and fixes could be given a rule like the following: After changing an implementation approach or fixing an issue, review the code logic and data definitions touched by the change, and make sure each one still has a reason to remain. Remove unused functions, redundant logic, and data fields that are no longer needed. When using a skill such as git-commit , the agent can automatically run this review before every commit, making sure unused functions, redundant logic, and data fields that are no longer needed are cleaned up promptly.

With AI, we need standards more, not less — otherwise we're just accelerating entropy. Ten projects, ten architectures, ten interaction patterns, ten data structures. Similar logic reinvented from scratch in each one.

A lot of AI products look like they are selling a product, but in practice they are often selling the model. From the customer’s point of view, delivery quality usually depends first on how strong the model is. So the real question becomes: if high-quality output mainly comes from the model, where does the product’s core competitiveness actually come from?

Thoughts do have force. If you keep leaning into the negative, a lot of the time you really do end up slowly pushing yourself further in that direction. The reverse is true too. You can start by almost tricking yourself a little, acting as if you are deeply curious, acting surprised by more foods and small things. Little by little, that curiosity really can grow back. It feels a bit like stamping a mental imprint into yourself, a gentle way of working on yourself from the inside.

Right now, a lot of AI is still landing at the efficiency stage. The market is also taking a harder look at whether that efficiency can really turn into returns, so sentiment keeps swinging. If the next step is finding a few more direct monetization points and letting the market see that AI can generate revenue, not just save costs, that probably feels like the closest catalyst for the next breakout phase of the AI theme.