引導式測驗 · 簡單 等級
學習使用人工智慧編碼作為初學者測驗
測試能力的錯覺、初學者使用人工智慧編碼助理的提示優先方式,以及為什麼必須檢查人工智慧編寫的程式碼。
問題 1 的 8
What does the guide call the core risk for beginners who let AI write their code?
預覽每個問題8 問題
當您參加測驗時,就會出現答案和解釋。
- What does the guide call the core risk for beginners who let AI write their code?
- Slower typing speed
- Higher electricity bills
- Losing access to the editor
- The illusion of competence
- Why do coding assistants sometimes invent functions that do not exist?
- They are programmed to test users
- They predict likely text rather than look up real, existing functions
- They only know very old programming languages
- Their internet connection drops
- What do editor-based assistants such as GitHub Copilot use as context for suggestions?
- Nearby code from your open files
- Your browsing history
- Only the programming language's official manual
- Nothing; every suggestion is random
- Which of these is a productive habit for beginners using AI?
- Paste the assignment and submit the output
- Only use AI after the deadline
- Try the problem first, then ask for hints rather than solutions
- Never run code the AI explains
- What was Harvard CS50's AI tutor, introduced in 2023, designed to do?
- Write complete solutions for every problem set
- Support learning without simply giving answers
- Grade exams automatically
- Replace all human teaching assistants
- Why is 'the code runs' not enough proof that AI-generated code is correct?
- Running code deletes it
- Code that runs is always slow
- Only teachers can run code
- It may still fail on unusual inputs or contain security flaws
- What configuration does the guide suggest while practising the basics?
- Turn off inline autocomplete
- Turn on maximum autocomplete
- Delete the chat history every minute
- Use two assistants at once
- What should you do before relying on a function the AI suggests?
- Ask the AI if it is sure
- Assume it exists if the name sounds right
- Read the documentation for it
- Rename it to something shorter