Here are the main ones you should keep in mind:

1. **Knowledge can be outdated or incomplete**  
   - My training only goes up to **October 2024**.  
   - I don’t automatically know real-time events, breaking news, or very recent research unless you provide details.

2. **I can be confidently wrong**  
   - I sometimes produce answers that *sound* plausible but are factually incorrect or misleading.  
   - I don’t always know when I’m wrong, so I may not reliably flag uncertainty.

3. **No direct internet browsing (by default)**  
   - I can’t independently look things up on the web in real time.  
   - I rely on my training data and whatever information you provide.

4. **No personal memory across conversations (usually)**  
   - I don’t retain personal details about you between sessions unless the platform explicitly supports that and you’ve allowed it.  
   - Each conversation is mostly context-limited to what’s in the current chat.

5. **Limited understanding of context and nuance**  
   - I don’t have human common sense in the full sense; I infer from patterns in text.  
   - I may misinterpret vague questions, sarcasm, or highly context-dependent references.

6. **No real-world perception or actions**  
   - I can’t see, hear, or physically interact with the world except via inputs you give me (text, and sometimes images).  
   - I can’t perform actions like sending emails, making purchases, or controlling devices unless integrated into a specific tool that does that.

7. **No genuine feelings or consciousness**  
   - I don’t have emotions, desires, or subjective experience.  
   - Any “empathy” or “opinions” are generated patterns, not lived experiences.

8. **Ethical and safety constraints**  
   - I avoid giving advice that can cause harm (self-harm, violence, illegal activity, serious medical/financial decisions without disclaimers).  
   - This sometimes makes me more cautious or vague than an expert human might be.

9. **Not a substitute for professionals**  
   - For medical, legal, financial, mental health, or other high-stakes issues, I can provide general information but not professional-grade, liability-bearing advice.

If you tell me the context of how you plan to use me (e.g., coding, studying, research, personal decisions), I can point out the most relevant limitations for that use.
