joheyhargrett / phase-1-ai-accelerate-your-learning-with-ai

phase-1 JavaScript

Accelerate Your Learning with AI

Learning Goals

• Describe the known-unknown matrix.
• Discuss guidelines on using AI for learning.

Introduction

In previous lessons, we discussed how AI chatbots are used by software engineers and the limitations of such tools. In this lesson, we will look at the known-unknown matrix and explore how we can leverage AI chatbots for learning programming concepts.

The Known-Unknown Matrix

The known-unknown matrix is a design thinking tool that helps to identify one’s familiarity with a topic.

	Known	Unknown
Known	Information or knowledge we think are true.	Information gaps or risks we are aware of.
Unknown	Information we know are true based on experiences but can’t explain.	Information we are not aware of at all.

Let’s look at examples for each of the quadrants to better understand the concept:

• Known-knowns: There are some fundamental programming principles you are aware of such as variable declarations and looping. You can use these ideas in your preferred programming language.
• Known-unknowns: Even if you exclusively use 1 or 2 programming languages, you may be aware of the existence of other programming languages. You know that they exist but aren’t aware of their data model or syntax.
• Unknown-knowns: You may use built-in or external functions and libraries in your application without fully understanding how they work. For example, you may use the sort method in JavaScript to sort an array without knowing what type of sorting algorithm is being used. You know from reading the docs and experience how to use the method but can’t explain how or why it works.
• Unknown-unknowns: These are things you don’t know exist. For example, if you have been only using programming languages that have loops, you may not be aware of languages that don’t have them (uses recursion instead).

How does labeling our knowledge help us effectively use AI chatbots? One of the main issues with these chatbots is their propensity to respond with completely false information with absolute certainty. If you’re unsure about your familiarity with a topic, chatbots may guide you in the wrong direction.

Here’s an exchange between a JS developer with 20+ years of experience and ChatGPT. In this case, Kyle asked a question that can be categorized as a known-known to him. He immediately knew that ChatGPT’s answer was wrong.

But if someone who wasn’t aware of the parseInt method and JavaScript’s type coercion rules asked this question, they may have taken ChatGPT’s answer for granted and miss the nuances. However, if they are aware of their lack of knowledge on the topic, they would seek out external sources for validation.

It’s important to know the limits of our knowledge so we know how to address them. A software engineer may have to work with multiple unfamiliar languages and tools at their job. This framework allows us to label our knowledge gaps, frame them in the context of our experiences, and then address them depending on the quadrant it’s in.

And finally, when you are learning something notice how your understanding of the topic is evolving. If you start at a specific stage and aren’t making progress, you should revise your learning strategy or seek out help.

Check out this article and this article for more on this subject.

Guidelines for Using AI Chatbots for Learning

In order to make the most of AI chatbots, we have to be aware of their limitations and structure our prompts accordingly. Here are some general guidelines that can help keep the chatbot from going off track:

1. Start Simple: You may encounter several complex prompts when researching online. But it’s completely fine to start with simple questions about concepts you want to learn more about and ask follow-up questions for further clarifications.
2. Code Examples: Chatbots can generate code samples along with line-by-line explanations for the code they generate. You can ask the chatbot to “provide examples to explain the concept of callbacks” as a starting point and then ask clarifications on specific parts of the code.
3. Real-world Examples: Chatbots can also provide case studies and show how concepts you may already know relate to new concepts you’re trying to learn. For example, when you’re learning about sorting, you can ask the chatbot, “what are some real-world applications of sorting algorithms?” and it will tell you about their role in search engines, databases, contact lists with detailed overview.
4. Code Tracing: Chatbots can provide a step-by-step walkthrough of code snippets along with explanations for each step. Provide the chatbot with your code snippet and ask it to “trace the execution path of the following code in ”.
5. Custom Quizzes: Custom quizzes can be generated on specific topics. Ask the chatbot, “give me a 3 question quiz challenge on JavaScript closures”. If it shows the answers along with the questions, you may have to add “don’t show the answers until I’ve tried to solve the questions” to the prompt.
6. Custom Code Challenges: We can also ask the chatbot to ask us a code challenge with the prompt, “give me a code challenge on closure in JavaScript”.
7. Validate with External Resources: If you are learning something and you’re not fully sure about the topic, validate the information provided by the chatbot with external sources. Ideal sources are official documentations, books, and high quality articles.

Custom Quiz Example

Here’s the user prompt:

Give me  a 3 question quiz challenge on JavaScript closures. Don't show the answers with the question.

And here’s the initial ChatGPT response with the questions:

Certainly! Here's a three-question quiz challenge on JavaScript closures. The answers will be provided separately after each question:

Question: What is the output of the following code snippet?

function outer() {
  var x = 10;
  function inner() {
    console.log(x);
  }
  return inner;
}

var closureFunc = outer();
closureFunc();

Question: Which of the following statements about closures in JavaScript is true?
	a) Closures are created when a function is defined.
	b) Closures only retain references to variables passed as arguments to inner functions.
	c) Closures cannot access variables from outer scopes.
	d) Closures are not used in modern JavaScript programming.

Question: In JavaScript, what happens if a closure references a variable that has been modified or reassigned after the closure was defined?
	a) The closure automatically updates to reflect the modified value.
	b) The closure retains the original value of the variable.
	c) The closure loses access to the variable and throws an error.
	d) The behavior depends on how the closure is invoked.

User prompt answering the questions:

Answer A: logs the value 10.
Answer B: A
Answer C: B

ChatGPT feedback response:

Great effort! Here are the correct answers:

Answer: The output of the code snippet will be 10.

Answer: The correct statement about closures in JavaScript is: a) Closures are created when a function is defined.

Answer: If a closure references a variable that has been modified or reassigned after the closure was defined, the closure retains the updated value. Therefore, your answer, option B, is correct.

Well done! If you have any more questions or need further clarification, feel free to ask!

Try asking it to create a code challenge in a similar manner. The challenges can be pretty fun!

Conclusion

AI chatbots are excellent tools for supporting education. General-purpose chatbots aren’t specifically designed for helping learners but they still offer powerful tools such as custom quiz generation and challenge generation, code walkthroughs, and code explanations.

Experiment with different ways of using a chatbot for learning. Occasionally reinforce your skills and knowledge through quizzes and challenges to ensure you’re retaining information. Remember that the guidelines are for helping you to get started but you can direct your own learning by leveraging these amazing chatbots!