引導式測驗 · 中等 等級
如何計算人工智慧專案測驗的投資報酬率
測試如何設定基準、調整人工智慧的採用率優勢、計算人工審核等隱藏成本,以及計算投資報酬率和投資回收期。
問題 1 的 8
Which formula matches the standard ROI calculation described in the guide?
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。
- Which formula matches the standard ROI calculation described in the guide?
- Total benefit divided by number of users
- (Total benefit minus total cost) divided by total cost
- Total cost divided by monthly benefit
- Time saved multiplied by salary
- How is the payback period calculated?
- Total benefit divided by total cost
- Number of months until adoption reaches 100 percent
- Upfront cost divided by monthly net benefit
- Monthly licence fee multiplied by 12
- A tool saves 30 minutes per user per day, but only 40 percent of staff use it regularly. What should the ROI model do?
- Scale the benefit down to reflect the 40 percent adoption rate
- Assume all staff will eventually use it and ignore the figure
- Double the benefit to account for enthusiastic users
- Exclude the tool from the analysis
- Which of these is an example of a hidden cost the guide says people often forget?
- The published monthly licence fee
- The vendor's list price per API call
- The cost of the office the team works in
- Staff time spent reviewing and correcting AI output
- According to the guide, when does time saved become real financial value?
- Automatically, as soon as the tool is deployed
- Only when freed hours are redeployed to valuable work, reduce overtime or avoid hiring
- Only when the vendor reports it in a dashboard
- Only if the AI runs overnight
- Why is measuring a baseline the first step?
- Without knowing current time, error rates and volume, any claimed improvement is a guess
- It is required by accounting law for AI projects
- It tells you which vendor to choose
- It replaces the need to measure results after launch
- How should error reduction be valued in an ROI model?
- By the number of AI features used
- By the accuracy score the vendor advertises
- By errors avoided multiplied by the cost per error
- By the percentage of tasks automated
- Why do pilot results often shrink when an AI tool is rolled out widely?
- Vendors deliberately slow tools down after pilots
- Larger teams always type more slowly
- ROI formulas change at scale
- Pilot users are often enthusiasts working on easier cases