quiz bu lëkkale ak tegtal · Diggante Niveau

IA ngir laaj ak tontu ci ndimbalu njiit yi ak njiit yi

Dafay saytu ni IA di soppi liggéeyu assistant executif: xool tukkib IA ak génne xibaar, yoriinu done yu nëbbu ci anam wu wóor, ak ni grounding di doxee.

Yooni njiit yu jëm ciIA ngir assistant executif
Laaj 1 de 8

Sunu sukkandikoo ci giid bi, lan mooy gëna am solo ginaaw bi IA jëlee liggéeyu assistant yu woyof yi?

Xoolaatal laaj bu nekk8 laaj

Sooy def quiz bi, tontu yi ak leeral yi dañuy feeñ.

  1. Sunu sukkandikoo ci giid bi, lan mooy gëna am solo ginaaw bi IA jëlee liggéeyu assistant yu woyof yi?
    • Typing speed and keyboard shortcuts
    • Judgment about what the executive needs, tone and confidentiality
    • Memorizing airline schedules
    • Manual transcription skills
  2. Su IA defar ab itinéraire tukki, lan la guide bi wax ni mooy liggéeyu assistant bi?
    • Booking immediately to save time
    • Asking the AI to rewrite the plan until it sounds confident
    • Forwarding the draft straight to the executive
    • Verifying schedules, visa rules and hotel details through authoritative sources
  3. Ban risk la tegtal bi di wax ci nettali yi IA defar ci benn moroom?
    • It can mix up people with similar names
    • It cannot read public biographies
    • It refuses to summarize news
    • It always leaves out recent events
  4. Lan la tegtal bi di wax ni mooy li gëna wóor ci mbir yu nëbbu yu melni këyitu tablo?
    • Any free consumer chatbot, since they are all similar
    • Pasting only half the document at a time
    • Enterprise versions with data protections, used within IT-approved rules
    • Emailing the material to a personal account first
  5. Ci tegtal bi, luy grounding?
    • Training a new model on all company email
    • Blocking AI from accessing the calendar
    • Retrieving items the user already has permission to see and adding them to the prompt
    • Printing AI outputs for manual review
  6. Why might an AI meeting summary attribute a decision to the wrong person?
    • Poor audio, crosstalk or missing speaker labels in the transcript
    • The meeting was too short to summarize
    • The model had no internet connection
    • Summaries are generated at random
  7. Which elements does the guide say a reliable prompt should include?
    • A single keyword
    • Humor to keep the model engaged
    • As much length as possible
    • A role, the context, the desired format and constraints
  8. As AI agents begin to act on an executive's behalf, what does the guide say a human should keep confirming?
    • Only font and formatting choices
    • Anything that spends money or commits the executive's time
    • Nothing, because agents are fully autonomous
    • Only internal memos