Start here: brand new to coding

Start with Python and small programs you understand. Your first win is making a script work, diagnosing why it breaks, and saving the change in Git. You can explore AI concepts alongside this without rushing into a complex application.

Your first week · 10 focused hours

1. 2 hours: start CS50P Functions, Variables. Type the examples and change the inputs. Watch the first lecture ↗.
2. 3 hours: complete a few exercises, then write a command-line expense or reading log. Use a list, a dictionary and a function.
3. 2 hours: make a CSV or JSON file reader. Deliberately give it a missing file and an invalid row; print a useful error.
4. 1 hour: follow GitHub Hello World. Save a README explaining how to run your script. Open the exercise ↗.
5. 2 hours: rebuild one tiny feature without generated code, explain it aloud, and write down what still confuses you.

The sequence I would follow

• First: CS50P with exercises. Prioritize units 0–6, then finish the remaining topics as they become useful. Follow the course’s own academic-honesty policy for submitted work.
• Next: selected CS50 SQL units on querying, relationships and designing tables; then Git and a small HTTP API.
• Then: the shared build path: one LLM feature → retrieval → evaluation → deployment → a domain project → applications.
• Planning allowance: 240–360 focused hours for foundations plus the shared path. At 10 hours/week, that is roughly 24–36 weeks before a first portfolio and interview cycle. This is my estimate, not a hiring timeline.

Use AI as a tutor

Ask for an explanation, a hint or a failing example before asking for a full solution. After help, close the answer and rebuild the idea. Keep a “bugs I fixed” page. If you can only make a change by pasting a whole file into an assistant, spend another session on that area.