AI tools: how to choose an AI coding assistant for your team

End-of-lesson check

15 questions · take it digitally for instant feedback at tendril.neural-forge.io/learn/quiz/end-tools-choosing-an-ai-coding-assistant-r7a1-creators

1. What are the three primary dimensions along which AI coding assistants vary, according to the material?

   1. Speed of response, accuracy percentage, and per-seat cost
   2. Team size requirements, project complexity, and deadline constraints
   3. Autonomy level, codebase awareness, and licensing terms
   4. Programming language support, plugin ecosystem, and documentation quality

2. What does 'autonomy level' refer to when evaluating AI coding assistants?

   1. The amount of storage space the tool requires on a developer's machine
   2. The number of different programming languages the tool can understand and assist with
   3. The speed at which the AI generates code completions in milliseconds
   4. The degree to which the AI can act independently to complete tasks without constant human oversight

3. Which of the following is listed as something AI coding assistants CAN do well?

   1. Replace the architectural judgment of senior engineers
   2. Tell you which mode fits your team's review culture
   3. Autocomplete inside the editor with low latency
   4. Guarantee that your code will never be used for model training

4. Before committing to an AI coding assistant, the lesson recommends which specific evaluation step?

   1. Reading every line of the tool's entire privacy policy in one sitting
   2. Asking the tool's support team which mode is best for your team
   3. Running a 2-week pilot on a real feature before committing
   4. Conducting a survey of other teams who use the tool

5. According to the material, what is a key security consideration when selecting an AI coding assistant?

   1. The number of employees who have access to your code
   2. The physical location of the company's servers
   3. Whether the tool sends your code to third parties for inference and what those parties do with the data
   4. Whether the tool requires an internet connection to function

6. Which evaluation criterion from the lesson directly addresses whether a tool might train its AI models on your proprietary code?

   1. Codebase awareness
   2. Per-seat cost
   3. License clarity
   4. Security model

7. A developer notices that one AI assistant only knows about the file currently open in their editor, while another has context about the entire repository including file relationships. What evaluation criterion captures this difference?

   1. Security model
   2. Codebase awareness
   3. Autonomy level
   4. License clarity

8. The lesson warns that you cannot rely on an AI coding assistant to do which of the following?

   1. Complete complex refactoring without introducing bugs
   2. Generate code that passes all automated tests on the first try
   3. Determine which autonomy mode fits your team's review culture
   4. Suggest the most efficient algorithm for a given problem

9. What does the lesson say about the relationship between a tool's underlying model and its suitability for your team?

   1. The choice matters more than which model is underneath
   2. The most popular model is the most suitable for teams
   3. The underlying model is the only factor that determines team fit
   4. The newest model is always the best choice for any team

10. A startup is comparing two AI coding assistants. Both have similar autonomy and awareness, but one costs $10/month per seat and the other costs $30/month per seat. Which evaluation criterion captures this difference?

    1. Per-seat cost
    2. Security model
    3. License clarity
    4. Autonomy level

11. Why does the lesson emphasize reading the 'subprocessor lists' of AI coding assistants?

    1. To understand which third parties might receive your code and for what purposes
    2. To check if the tool can integrate with your version control system
    3. To determine the maximum file size the tool can process
    4. To verify that the tool supports your preferred programming languages

12. A team lead is trying to decide between three AI assistants: one operates as an autocomplete, one as a chat-based assistant, and one as a full agent. What should the team lead primarily consider when choosing among these?

    1. Which assistant has the longest privacy policy
    2. Which assistant has the most attractive user interface
    3. Which assistant was released most recently
    4. How much independent action they want the AI to have and their team's review process

13. What does the lesson identify as a common misconception about AI coding assistants?

    1. That AI assistants work better than human code reviewers
    2. That AI assistants require no internet connection to function
    3. That AI assistants can write code faster than human developers
    4. That the hottest or most popular tool is always the right tool for any team

14. When the lesson mentions 'security model,' which of the following is NOT typically part of this evaluation criterion?

    1. Whether the tool sends code to the cloud for processing
    2. Whether the tool can be deployed in a VPC environment
    3. Whether the tool can run entirely locally
    4. The specific autonomy mode the tool operates in

15. A developer wants to evaluate whether an AI coding assistant might use their code to train future AI models. Which combination of evaluation criteria would be most relevant?

    1. Per-seat cost and pilot testing
    2. Codebase awareness and autonomy level
    3. License clarity and security model
    4. Autonomy level and codebase awareness