The article discusses how AI can be trained to assist developers in coding, likening the AI to an intern with potential but lacking judgment.It emphasizes the importance of mentoring and guiding AI like an apprentice to improve its coding skills.The concept of 'AI Pair Programming' is introduced as a way to work with AI in a structured manner to solve coding problems.Guidelines are provided on setting up an AI coding flow, outlining problems, coding basic versions, and documenting the process.The article suggests training AI to learn developer preferences and habits to produce cleaner code with fewer bugs.Building feedback loops for auto-evaluation is recommended to help AI improve itself based on set standards and unit tests.Resources like Typer Docs, Rich, Python Type Hints, and pytest are suggested to aid in smart prompting and debugging while training AI.The article concludes by emphasizing that developers are not being replaced by AI but rather elevated through the mentorship and training of AI tools.It encourages developers to start training AI using scripts, guiding them through each project and prompt, with a reminder to come back for support.Additionally, ten recommended courses related to computing, tech history, and programming weirdness are highlighted for further learning.