引導式測驗 · 簡單 等級
工人人工智慧素養測驗
測試工作場所人工智慧素養的核心能力:模型為何產生幻覺、如何驗證輸出以及根據歐盟人工智慧法案仔細處理資料。
問題 1 的 8
根據指南,職場人工智慧素養是否需要學習程式設計?
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。
- 根據指南,職場人工智慧素養是否需要學習程式設計?
- Yes, Python is the first requirement
- No, it is a set of working habits for using AI well and safely
- Only for managers
- Only in the European Union
- 什麼是生成人工智慧的幻覺?
- A deliberate lie programmed by the vendor
- An image with distorted colors
- A confident statement from the model that is false
- A tool crash caused by a long prompt
- 為什麼模型可以發明看起來完全真實的引文?
- It generates likely text from patterns and has no built-in step that checks claims against a database
- It copies citations from a secret list
- It is designed to test users
- Citations are always generated by a separate human team
- 歐盟人工智慧法案第 4 條(人工智慧素養義務)從什麼時候開始適用?
- 1 January 2020
- 1 August 2023
- 31 December 2027
- 2 February 2025
- 範例中的行銷協調員在使用未經批准的聊天機器人之前做了什麼?
- Asked the customer for permission to share their data
- Removed the customer's name, account number and order history
- Switched off his internet connection
- Pasted the email in full and deleted it afterward
- Does retrieval-augmented generation (RAG) eliminate hallucinations?
- Yes, completely
- Yes, but only for images
- No, it reduces them, but the model can still misread or misattribute retrieved documents
- No, it makes them more frequent in every case
- Why did the new hire's question about parental leave get only a generic answer?
- The chatbot had no access to the company's internal documents
- Parental leave is a banned topic for AI
- The chatbot was deliberately withholding information
- The question was too short
- According to the guide, who is responsible for work that uses AI output?
- 人工智慧供應商
- The IT department
- No one, if the AI wrote it
- The person who uses the output