What AI hallucinations are in AI for Everyone - Mini Project: Build & Apply

Understanding What AI Hallucinations Are

📖 Scenario: You are learning about artificial intelligence (AI) and how it sometimes makes mistakes called hallucinations. These mistakes happen when AI gives answers that sound real but are actually wrong or made up.

🎯 Goal: Build a simple explanation and example list that shows what AI hallucinations are and how to recognize them.

📋 What You'll Learn

Create a dictionary called ai_hallucinations with three examples of AI hallucinations and their explanations.

Add a variable called example_count that stores the number of examples in the dictionary.

Use a for loop with variables example and description to iterate over ai_hallucinations.items().

Add a final statement that sets a variable summary to a short explanation of AI hallucinations.

💡 Why This Matters

🌍 Real World

Understanding AI hallucinations helps people use AI tools more carefully and avoid trusting wrong information.

💼 Career

Knowledge of AI hallucinations is important for jobs in AI development, content creation, and digital literacy education.

1

Create the AI hallucinations dictionary

Create a dictionary called ai_hallucinations with these exact entries: 'Wrong facts' with value 'AI gives incorrect information that sounds true.', 'Made-up sources' with value 'AI invents references or quotes that do not exist.', and 'Confused context' with value 'AI mixes up details from different topics.'.

AI for Everyone

# Create the ai_hallucinations dictionary with 3 examples
# Your code here

Hint

Use curly braces {} to create a dictionary with keys and values as shown.

2

Add a variable for the number of examples

Add a variable called example_count that stores the number of entries in the ai_hallucinations dictionary using the len() function.

AI for Everyone

ai_hallucinations = {
    'Wrong facts': 'AI gives incorrect information that sounds true.',
    'Made-up sources': 'AI invents references or quotes that do not exist.',
    'Confused context': 'AI mixes up details from different topics.'
}
# Add example_count variable here
# Your code here

Hint

Use len(ai_hallucinations) to get the number of items in the dictionary.

3

Loop through the AI hallucination examples

Use a for loop with variables example and description to iterate over ai_hallucinations.items(). Inside the loop, write a comment explaining that this loop goes through each hallucination example and its explanation.

AI for Everyone

ai_hallucinations = {
    'Wrong facts': 'AI gives incorrect information that sounds true.',
    'Made-up sources': 'AI invents references or quotes that do not exist.',
    'Confused context': 'AI mixes up details from different topics.'
}
example_count = len(ai_hallucinations)
# Use a for loop to go through each example and description
# Your code here

Hint

Use for example, description in ai_hallucinations.items(): to loop through the dictionary.

4

Add a summary explanation variable

Add a variable called summary and set it to the string 'AI hallucinations are mistakes where AI gives false or made-up information.'.

AI for Everyone

ai_hallucinations = {
    'Wrong facts': 'AI gives incorrect information that sounds true.',
    'Made-up sources': 'AI invents references or quotes that do not exist.',
    'Confused context': 'AI mixes up details from different topics.'
}
example_count = len(ai_hallucinations)
for example, description in ai_hallucinations.items():
    # This loop goes through each hallucination example and its explanation
    pass
# Add the summary variable below
# Your code here

Hint

Set summary to the exact string given.

Practice

1. What does the term AI hallucination mean?

easy

A. An AI giving a confident but incorrect or made-up answer

B. An AI dreaming during processing

C. An AI shutting down unexpectedly

D. An AI learning from new data

Solution

1. Step 1: Understand the meaning of AI hallucination

   AI hallucination refers to when an AI produces an answer that sounds confident but is actually wrong or made up.

2. Step 2: Match the definition to the options

   An AI giving a confident but incorrect or made-up answer correctly describes this as a confident but incorrect or made-up answer, while other options describe unrelated AI behaviors.

3. Final Answer:

   An AI giving a confident but incorrect or made-up answer -> Option A

4. Quick Check:

   AI hallucination = wrong confident answer [OK]

Hint: Remember: hallucination means wrong but confident answer [OK]

Common Mistakes:

• Thinking AI hallucination means AI dreaming
• Confusing hallucination with AI failure or crash
• Assuming hallucination means learning new facts

2. Which of the following best describes a sign of an AI hallucination?

easy

A. The AI refuses to answer any question

B. The AI repeats the same answer multiple times

C. The AI asks for more information

D. The AI gives an answer that sounds sure but is false

Solution

1. Step 1: Identify the sign of AI hallucination

   AI hallucinations happen when the AI gives answers that sound confident but are actually false or made up.

2. Step 2: Compare options to this definition

   The AI gives an answer that sounds sure but is false matches this description exactly. Other options describe different AI behaviors not related to hallucinations.

3. Final Answer:

   The AI gives an answer that sounds sure but is false -> Option D

4. Quick Check:

   Hallucination = confident but false answer [OK]

Hint: Look for confident but false answers as hallucinations [OK]

Common Mistakes:

• Confusing refusal to answer with hallucination
• Thinking repeated answers are hallucinations
• Assuming asking for info means hallucination

3. Consider this AI response to a question: "The capital of Mars is Olympus City." What is this an example of?

medium

A. Correct factual answer

B. AI asking for clarification

C. AI hallucination

D. AI refusing to answer

Solution

1. Step 1: Analyze the AI response content

   The AI claims a capital city on Mars, which is a fictional or made-up fact since Mars has no capital.

2. Step 2: Identify this as an AI hallucination

   Because the AI confidently gives a false or made-up answer, this fits the definition of an AI hallucination.

3. Final Answer:

   AI hallucination -> Option C

4. Quick Check:

   Made-up confident answer = hallucination [OK]

Hint: False confident facts from AI = hallucinations [OK]

Common Mistakes:

• Assuming fictional facts are correct
• Thinking AI refused to answer
• Confusing hallucination with clarification requests

4. An AI gives this answer: "The Eiffel Tower is located in Berlin." What is the best way to fix this hallucination?

medium

A. Check facts and correct the location to Paris

B. Ignore it because AI is always right

C. Restart the AI to fix the error

D. Ask the AI to repeat the same answer

Solution

1. Step 1: Identify the error in the AI answer

   The AI incorrectly states the Eiffel Tower is in Berlin, which is false; it is in Paris.

2. Step 2: Choose the best fix for the hallucination

   The best fix is to check the facts and correct the answer to Paris, ensuring accuracy.

3. Final Answer:

   Check facts and correct the location to Paris -> Option A

4. Quick Check:

   Fix hallucination by fact-checking and correcting [OK]

Hint: Always verify AI answers to fix hallucinations [OK]

Common Mistakes:

• Ignoring wrong AI answers
• Restarting AI without fact-checking
• Repeating wrong answers expecting change

5. You receive this AI-generated summary: "The Great Wall of China was built in the 20th century to protect against alien invasions." How should you handle this hallucination when using the summary for a school project?

hard

A. Use it as is because AI sounds confident

B. Cross-check with reliable sources and correct the facts

C. Ignore all AI summaries for school projects

D. Add more made-up details to make it interesting

Solution

1. Step 1: Recognize the hallucination in the summary

   The claim about the Great Wall being built in the 20th century for alien invasions is false and made up.

2. Step 2: Decide the best approach for using AI content

   Always cross-check AI-generated facts with trusted sources and correct any errors before use.

3. Final Answer:

   Cross-check with reliable sources and correct the facts -> Option B

4. Quick Check:

   Verify AI info before use to avoid hallucination errors [OK]

Hint: Always verify AI info before using it [OK]

Common Mistakes:

• Trusting AI without fact-checking
• Discarding all AI content unnecessarily
• Adding false info to cover errors