> ## Documentation Index
> Fetch the complete documentation index at: https://docs.omnifact.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# AI Capabilities and Limitations

> Understanding what AI can and cannot do to set realistic expectations

export const Glossary = ({children, term}) => {
  const glossaryDefinitions = {
    'ai': 'Artificial Intelligence - computer systems that can perform tasks typically requiring human intelligence',
    'artificial-intelligence-ai': 'Computer systems that can perform tasks typically requiring human intelligence',
    'ai-assistant': 'An AI system designed to help users with tasks through conversation',
    'space': 'A dedicated environment with a specialized AI assistant and optional Uploaded Files',
    'spaces': 'Dedicated environments with specialized AI assistants and optional Uploaded Files',
    'knowledge': 'The sub-menu in Space settings where you manage Uploaded Files and Connected Sources',
    'uploaded-files': 'Documents manually uploaded to a Space under the Knowledge tab',
    'prompt': 'Instructions or questions you give to an AI assistant',
    'response': 'The AI assistant\'s answer to your prompt or question',
    'query': 'A question or request for information',
    'context': 'Background information that helps the AI understand your request',
    'llm': 'Large Language Model - an AI system trained on vast amounts of text data to understand and generate human language',
    'large-language-model-llm': 'An AI system trained on vast amounts of text data to understand and generate human language',
    'privacy-filter': 'A system that automatically detects and masks sensitive information before it reaches AI models, and restricts images from being sent to non-EU hosted models for compliance',
    'chat-instructions': 'Custom guidelines that define how an AI assistant behaves and responds in a Space',
    'pattern-matching': 'The process by which AI recognizes and uses patterns in data to provide responses',
    'patterns': 'Recognizable structures or trends in data that AI systems can identify and use',
    'training-data': 'The text and information used to teach an AI system how to understand and respond to requests',
    'natural-language': 'Everyday human language that people use to communicate, as opposed to computer code or formal syntax',
    'hallucination': 'When an AI generates information that sounds plausible but is actually incorrect or fabricated',
    'filtering': 'The process of automatically detecting and temporarily replacing sensitive information to protect it during AI processing',
    'masking': 'Temporarily replacing sensitive information with generic labels to protect it from being seen by external AI models',
    're-insertion': 'The automatic process of putting original sensitive information back into AI responses after they\'ve been generated with filtered data',
    'retrieval': 'The process of finding and accessing relevant documents from Uploaded Files to answer questions',
    'hosting': 'Where an AI model is deployed and made available for use - either by third-party providers, on your organization\'s own servers, or in private cloud environments',
    'cite': 'To reference or mention the source of information, helping you know where an AI\'s answer came from',
    'sidebar': 'A panel that appears on the side of the screen showing additional information and options',
    'toggle': 'A button or control that switches something on or off, like opening and closing a sidebar',
    'navigation': 'The system of menus and buttons that help you move between different parts of Omnifact',
    'favoriting': 'Marking a Space as a favorite to pin it to the main navigation for quick access',
    'pinning': 'Attaching a Space to the main navigation bar so it\'s always visible and easily accessible'
  };
  const definition = glossaryDefinitions[term];
  return <Tooltip tip={definition}>{children}</Tooltip>;
};

## Know Your AI Partner

To get the most out of Omnifact, it helps to know what it's good at—and where it needs your help. Think of AI as a brilliant but literal-minded intern: incredibly fast and knowledgeable, but lacking real-world experience.

<CardGroup cols={2}>
  <Card title="Superpower" icon="bolt">
    Processing massive amounts of data instantly.
  </Card>

  <Card title="Kryptonite" icon="link-slash">
    Understanding nuance, sarcasm, and complex human context.
  </Card>
</CardGroup>

## What AI Does Best

<Glossary term="ai">AI</Glossary> excels at several key areas that can transform your everyday work. It can rapidly **process and analyze** large volumes of information, spot important **<Glossary term="patterns">patterns</Glossary> and trends** in data, **generate high-quality professional content**, and provide **detailed answers** to your questions. Think of it as your own personal assistant that combines the speed of a computer with the communication abilities of a skilled professional.

## Real-World Success Examples

<Tabs>
  <Tab title="Document Analysis" icon="file-lines">
    **Task Example:** Analyzing customer feedback reports

    **What AI Can Do:**

    * Summarize 100+ feedback responses in minutes
    * Identify common themes and issues
    * Track sentiment trends over time
    * Suggest areas for improvement

    **Example Request:**
    "Analyze these 150 customer feedback responses about our new feature. Summarize the top 3 issues mentioned, track sentiment changes over the past month, and suggest two areas for immediate improvement."
  </Tab>

  <Tab title="Content Creation" icon="pen-to-square">
    **Task Example:** Creating professional, multilingual communications

    **What AI Can Do:**

    * Draft clear, well-structured emails, documents, and presentations
    * Edit and improve existing content for tone, clarity, and grammar
    * Translate content between languages while maintaining tone and context
    * Adapt technical content for different audiences or regions

    **Example Request:**
    "Help me write a professional email announcing our team's successful project completion, highlighting our key achievements and thanking contributors. Then translate it into French, preserving the tone and formality."
  </Tab>

  <Tab title="Research Support" icon="books">
    **Task Example:** Market research analysis

    **What AI Can Do:**

    * Compare competitor features across multiple products
    * Identify emerging market trends from various sources
    * Summarize lengthy industry reports into actionable insights
    * Generate data-driven recommendations for strategy

    **Example Request:**
    "Analyze these three industry reports from the past year. Identify the top 5 emerging trends, compare our product features to our main competitors, and suggest three potential areas for innovation in our product line."
  </Tab>

  <Tab title="Process Improvement" icon="gears">
    **Task Example:** Workflow optimization

    **What AI Can Do:**

    * Document complex procedures in clear, step-by-step formats
    * Suggest efficiency improvements based on best practices
    * Create comprehensive checklists for critical processes
    * Streamline and simplify technical instructions

    **Example Request:**
    "Review our current employee onboarding process. Create a streamlined, step-by-step guide, suggest three ways to improve efficiency, and develop a checklist for HR to ensure all steps are completed."
  </Tab>

  <Tab title="Data Analysis" icon="chart-line">
    **Task Example:** Financial data interpretation

    **What AI Can Do:**

    * Analyze large, complex datasets quickly and accurately
    * Identify subtle patterns and anomalies in financial data
    * Generate visual representations of data for easy understanding
    * Provide predictive insights based on historical trends

    **Example Request:**
    "Analyze our sales data for the past year. Create a report showing monthly trends, highlight any significant changes or patterns, and generate a forecast for the next quarter based on these trends. Include visual charts for key metrics."
  </Tab>
</Tabs>

## Important Limitations

<CardGroup cols={2}>
  <Card title="No True Understanding" icon="brain">
    AI matches patterns but doesn't truly comprehend like humans
  </Card>

  <Card title="Knowledge Has Limits" icon="graduation-cap">
    Base knowledge comes from training data, but Omnifact can also search the web, your Uploaded Files, and attached documents when those capabilities are enabled
  </Card>

  <Card title="No Real Experience" icon="user-xmark">
    Lacks real-world experience and common sense
  </Card>

  <Card title="No Emotional Intelligence" icon="heart">
    Cannot truly understand or respond to emotions
  </Card>
</CardGroup>

<Note>
  Understanding AI's limitations helps you use it more effectively and know when to rely on human expertise.
</Note>

### The Confidence Trap: AI as a "People Pleaser"

AI models are designed to be helpful, which can sometimes make them "people pleasers." They will try to give you an answer even if they have to invent one, rather than saying "I don't know."

**Crucial Warning:** The AI sounds **exactly the same** when it is telling the truth as when it is hallucinating (making things up). It does not sound uncertain.

<Warning>
  **The Rule of Thumb:** The more professional and confident the AI sounds, the more you need to check the facts. Never assume a confident tone equals accuracy.
</Warning>

## Specific Examples of Limitations

<Tabs>
  <Tab title="Complex Decision Making" icon="scale-balanced">
    **What AI Cannot Do:**

    * Make final business or ethical decisions
    * Evaluate complex moral dilemmas
    * Judge subjective or situational matters
    * Handle sensitive personnel issues

    **Example:**\
    AI can provide data and analysis about different strategies, but it shouldn't make final decisions—especially when ethical tradeoffs are involved. For instance, AI can't decide whether to enter a new market based on cultural factors, long-term brand impact, or potential environmental consequences that require weighing short-term profits against sustainability.
  </Tab>

  <Tab title="Creative Strategy" icon="lightbulb">
    **What AI Cannot Do:**

    * Develop original business strategies
    * Create innovative solutions
    * Understand market nuances
    * Make intuitive leaps

    **Example:** AI can help brainstorm ideas but shouldn't determine your company's long-term vision or strategic direction. It can't, for example, decide on a new product line based on emerging cultural trends and potential future customer needs.
  </Tab>

  <Tab title="Personal Interactions" icon="users">
    **What AI Cannot Do:**

    * Handle sensitive employee matters
    * Resolve conflicts
    * Build genuine relationships
    * Show true empathy

    **Example:** While AI can help draft communication, it shouldn't handle sensitive conversations like performance reviews or conflict resolution. For instance, AI can't effectively mediate a dispute between team members with complex interpersonal dynamics.
  </Tab>

  <Tab title="Contextual Understanding" icon="sitemap">
    **What AI Cannot Do:**

    * Fully grasp cultural nuances
    * Understand implicit social cues
    * Interpret sarcasm or humor reliably
    * Adapt to rapidly changing situations

    **Example:** AI might misinterpret culturally specific jokes or fail to recognize when a statement is meant sarcastically, potentially leading to miscommunication in international business contexts.
  </Tab>
</Tabs>

## Best Practices for Working with AI

<Card title="Verify Important Information" icon="check-double" horizontal>
  Always fact-check critical information from AI responses
</Card>

<Card title="Use Human Judgment" icon="brain" horizontal>
  Combine AI suggestions with your expertise and experience
</Card>

<Card title="Set Clear Boundaries" icon="border-all" horizontal>
  Know which tasks to delegate to AI and which require human handling
</Card>

<Card title="Maintain Oversight" icon="eye" horizontal>
  Review and validate AI outputs before using them
</Card>

<Warning>
  AI can sometimes <Glossary term="hallucination">hallucinate</Glossary> or generate incorrect information that sounds convincing. Always fact-check important information, especially for business-critical decisions. Don't assume AI responses are accurate just because they sound authoritative - verify key facts from reliable sources.
</Warning>

## When to Use AI vs. Human Expertise

<CardGroup cols={2}>
  <Card title="🤖 Great for AI">
    • Document analysis and summarization (reports, research papers, articles) <br />
    • Creating initial content drafts (emails, reports, documentation) <br />
    • Processing and analyzing large datasets <br />
    • Research support and information gathering <br />
    • Repetitive task automation <br />
    • Pattern recognition in data <br />
    • Language translation and proofreading <br />
    • Basic question answering <br />
  </Card>

  <Card title="👤 Needs Human Touch">
    • Final business and strategic decisions <br />
    • Long-term strategic planning and vision setting <br />
    • Sensitive personnel matters and HR issues <br />
    • Creative direction and innovative thinking <br />
    • Complex ethical decisions <br />
    • Building authentic relationships <br />
    • Emotional intelligence and empathy <br />
    • Critical judgment calls <br />
    • Conflict resolution and mediation <br />
    • High-stakes negotiations <br />
  </Card>
</CardGroup>

## Tips for Success

<AccordionGroup>
  <Accordion title="Getting Better Results" icon="arrow-up-right">
    * Be specific in your requests
    * Provide clear context
    * Break down complex tasks
    * Review and refine outputs
    * Use follow-up questions
  </Accordion>

  <Accordion title="Avoiding Common Pitfalls" icon="triangle-exclamation">
    * Don't rely on AI for final decisions
    * Always verify critical information
    * Keep sensitive matters private
    * Maintain human oversight
    * Use appropriate security settings
  </Accordion>

  <Accordion title="Maximizing Value" icon="chart-line">
    * Start with simple tasks
    * Build complexity gradually
    * Learn from experience
    * Share best practices
    * Document successful approaches
  </Accordion>
</AccordionGroup>

<Tip>
  AI is a powerful tool that works best when combined with human expertise. Use it to enhance your work, not replace your judgment.
</Tip>

## Next Steps

Now that you understand AI's capabilities and limitations, explore these topics to learn more:

* Learn about [How to Interact with AI](/en/guides/ai-fundamentals/how-to-interact) to communicate effectively
* Check the [Glossary](/en/guides/ai-fundamentals/glossary) for explanations of any terms you've encountered
* Review [Understanding AI Assistants](/en/guides/ai-fundamentals/understanding-assistants) to see how AI can help your work