Home > Information Technology > Artificial Intelligence Roadmap > AI Resource Guide

AI Resource Guide

Generative AI is a type of artificial intelligence that can learn from and mimic large amounts of data to create content such as text, images, music, videos, code, and more, based on inputs or prompts. The County supports responsible use and experimentation with generative AI, but there are important considerations to keep in mind while using these tools:

  • Protect confidential data: Avoid inputting sensitive data into publicly available AI tools. This may lead to unauthorized individuals gaining access to that data.
  • Review content before publication: AI-generated content can be inaccurate, misleading, or entirely fabricated. It may also contain copyrighted material. Evaluate all AI-generated material before releasing it to the public.
  • Adhere to existing County policy: Comply with all county and department policies regarding data handling and confidentiality. View Miami Dade County’s AI Policy for more information.
  •  
    Tool Name Strengths/Limitations Ethical Considerations
    ChatGPT*
    • Strength: Excellent for conversational tasks, creative writing, and brainstorming
    • Limitation: Prone to inaccuracies and hallucinations due to lack of real-time data retrieval
    • Bias and Discrimination: Can replicate biases present in its training data
    • Data Privacy: May process sensitive personal information
    • Transparency: Potential for incorrect or misleading information
    • Misuse Potential: Risks of creating deceptive content or plagiarism
    Gemini*
    • Strength: Advanced personalized learning capabilities and multilingual support
    • Limitation: Challenges in identifying subtle biases and addressing misinformation
    • Bias: Outputs may reflect biases in training data
    • Transparency: Difficult to understand decision-making process
    • Misinformation: Can be used to create deceptive content
    Copilot for Microsoft 365*
    • Strength: Integrates with Microsoft tools to enhance productivity
    • Limitation: Limited functionality outside of Microsoft ecosystem
    • Privacy: User data handling and confidentiality
    • Transparency: Clear understanding of capabilities and limitations
    • Bias and Fairness: Efforts to minimize biases
    Perplexity*
    • Strength: Offers detailed citations and contextual research
    • Limitation: Susceptible to algorithmic bias
    • Data Bias: Outputs may reflect biased training data
    • Algorithmic Bias: Potential skewing in results
    Claude*
    • Strength: Strong ethical safeguards and clear explanations
    • Limitation: Limited availability and less suited for complex creative tasks
    • Bias and Discrimination: Outputs may reflect societal biases
    • Data Privacy: Handles sensitive information securely