Hi, Have you created a plan for the AI to code for 24/7 and also have you created agents to test and validate the code, that way the issues in the code can be mitigated mostly
It seems that a significant number of developers may be using AI agents at their own expense, as companies tend to provide access primarily when building projects around it — often with the goal of reducing personnel costs and time. This is, of course, an observation and may vary across different organizations and industries.
Based on my observations, creating a directory with relevant claude.md files tends to increase file count, and claude --resume already stores code-related data. Given this, I'm wondering if it might be worth considering an alternative approach — summarizing the entire code and its actions instead. From what I understand, Claude handles this through a planning process, where we define a to-do list before starting a project or idea, iterate on it until we're satisfied, and then ask it to build accordingly. With this approach in place, I'm curious whether HAM is still necessary — just wanted to share this thought for consideration, not to dismiss anything.
I think the point mentioned in the Substack was spot on, especially the example of calculator. Just as calculators did not eliminate our ability to understand math but helped us work more efficiently, AI will not take away our basic skills if we focus on proper understanding. Communication and critical thinking should be given more priority than simply relying on knowledge from books and tools