Specify the Output

Common Output Formats

• Bullet points: useful for lists of features, steps, or key points
• Numbered list: good for ordered steps or ranked items
• Paragraph: good for explanations and narrative content
• Table: good for comparisons and structured data
• Code block: essential for any programming output
• JSON: useful for structured data that needs to be used programmatically
• Markdown: useful for content that will be used in documentation

Output Format in Practice

# How output format changes the response

# Without format specification:
no_format = "Compare Python and JavaScript."
# AI may write paragraphs, a list, or a table.
# You get whatever format AI prefers.

# With format specification:
with_format = """
Compare Python and JavaScript using a table with 4 rows.
Rows: Use case, Learning curve, Performance, Community support.
Use simple language. One cell per row per language.
"""
# AI produces exactly the table you need.

# Another example:
step_format = """
Explain how to install Python on Windows.
Use a numbered list of steps.
Start from downloading the installer.
End at verifying the installation in the terminal.
"""
# AI produces a clear numbered guide.

print("Output format = what shape the response takes.")

Key Takeaways

• Specifying output format prevents AI from choosing a format that does not suit your needs
• Common formats: bullet points, numbered list, table, paragraphs, code, JSON
• Format choice depends on how you will use the output
• You can combine formats: 'Write an introduction paragraph followed by 3 bullet points'
• If format matters, always specify it explicitly