The Official ChatGTP cheat code thread.
- Thread starter Ethnic Vagina Finder
- Start date
null
...
It’s possible. I build my first software applicationworking on my second one now. It’s not easy tho. Developers shouldn’t fear people taking their jobs.
not all animals (apps) are equal.
not all animals (apps) are equal.
can you provide more details into the issues you've come across?
Ethnic Vagina Finder
The Great Paper Chaser
Another thing I learned is sometimes, it's not a good idea to keep asking the AI to make changes or revisions. Because eventually, it will start fukking up.
And ALL AI platforms will eventually take you on an error loop. Meaning, you get an error, it gives you Solution 1. You get the same error and it gives you Solution 2. You get a new error and it gives you Solution 1 all over again, taking you back to the first error.
It will also give you solutions, that are not in the official documentation. I'm using Wails and Svelte. Frameworks and platforms a small ecosystem is more challenging to work with using AI because they don't have enough information stored to give you accurate solutions.
Last edited:
1/1
@MicahBerkley
Free Game - Using Claude to Modify x.com
I'm taking @RayFernando1337's advice... but let me show you an actual perfect use case of creating JS code or bookmarklet to do something the site wasn't designed to do.
I don't have time to uncheck all 479 interests that are turned on for my account. So I goto Claude have it create a bookmarklet or JS code to do this function for me.
This is so clutch. I gave Claude a sample of code from the page using the inspect function. Copied the HTML so it knew what it was dealing with and it wrote the code almost flawlessly.
Let me know if I should dive deeper on this.
[Quoted tweet]
Here is how you fix your timeline. Go to:
Settings & Privacy -> Privacy and safety -> Content you see -> unfollow topics and uncheck all interests
On web the setting is in:
Settings & Privacy -> Privacy and safety -> Ads Preferences -> Uncheck all interests
https://video.twimg.com/ext_tw_video/1823008334976327680/pu/vid/avc1/938x720/CUDwmIWpOh3j7jjh.mp4
To post tweets in this format, more info here: https://www.thecoli.com/threads/tips-and-tricks-for-posting-the-coli-megathread.984734/post-52211196
@MicahBerkley
Free Game - Using Claude to Modify x.com I'm taking @RayFernando1337's advice... but let me show you an actual perfect use case of creating JS code or bookmarklet to do something the site wasn't designed to do.
I don't have time to uncheck all 479 interests that are turned on for my account. So I goto Claude have it create a bookmarklet or JS code to do this function for me.
This is so clutch. I gave Claude a sample of code from the page using the inspect function. Copied the HTML so it knew what it was dealing with and it wrote the code almost flawlessly.
Let me know if I should dive deeper on this.
[Quoted tweet]
Here is how you fix your timeline. Go to:
Settings & Privacy -> Privacy and safety -> Content you see -> unfollow topics and uncheck all interests
On web the setting is in:
Settings & Privacy -> Privacy and safety -> Ads Preferences -> Uncheck all interests
https://video.twimg.com/ext_tw_video/1823008334976327680/pu/vid/avc1/938x720/CUDwmIWpOh3j7jjh.mp4
To post tweets in this format, more info here: https://www.thecoli.com/threads/tips-and-tricks-for-posting-the-coli-megathread.984734/post-52211196
Meta Prompting with o1, o1 Pro Mode, and ChatGPT Pro (Compute on Compute)
IndyDevDan
35.1K subscribers
38,564 views Dec 9, 2024 #promptengineering #llm #aiengineering
What is meta-prompting and is O1 Pro Mode just a flashy CASH-grab? Metaprompting is a powerful advanced prompt engineering technique and No, o1 Pro Mode on ChatGPT Pro is an incredible GenAI tool that you can use to supercharge your productivity and get more out of your COMPUTE. The big question is... Should you pay $200/month for it and can you REALLY use the reasoning compute? 
VIDEO REFERENCES • GET MY METAPROMPT: https://gist.github.com/disler/29ff18... • CHATGPT PRO: https://openai.com/index/introducing-... • COMPLETE O1: https://openai.com/index/openai-o1-sy... • HN DISCUSSION: https://news.ycombinator.com/item?id=... • PREVIOUS QWQ PROMPT CHAINING VIDEO: 
• Yup, QwQ is CRACKED: Prompt Chaining ... • PROMPT FORMAT VIDEO (XML):
• BEST Prompt Format: Markdown, XML, or... In this video, we put O1, and O1 Pro Mode, in ChatGPT Pro head-to-head to see if paying premium is truly worth it. Learn how meta prompting and the metaprompt technique can transform your entire approach to AI-driven productivity. We break down the power of these cutting-edge reasoning models so you can decide if that steep monthly fee aligns with your workflow and ROI. We explore how meta prompting and the metaprompt gives you assymetric productivity in the age of Generative AI. Watch as we pit O1 vs O1 Pro Mode and discover if ChatGPT Pro can justify its hefty price tag. We’ll break down how these advanced reasoning models from OpenAI can supercharge your prompt engineering and push beyond standard large language model outputs. Whether you’re a seasoned developer or just starting out, we’ll show you how to harness meta prompting for maximum impact across a wide range of prompts. Ready to refine your prompt engineering skills, streamline your agentic workflows, and tap into next-level capabilities? This video is your ticket to unlocking O1, O1 Pro Mode, ChatGPT Pro, and the entire reasoning model revolution. 
What you'll learn:- How meta prompting can supercharge your AI workflows
- Real-world comparisons between O1 and O1 Pro Mode
- Whether ChatGPT Pro's $200 price tag is justified for your needs
- Advanced techniques for maximizing LLM performance
- The future of compute-intensive AI features
Key findings: O1 Pro Mode shows a 0-20% performance improvement over base O1, but is this marginal gain worth the investment? For most engineers, the base O1 model will suffice - but power users pushing the boundaries of compute reasoning will find significant value in Pro Mode's enhanced capabilities. 
Pro Tip: The prompt is the fundamental unit of knowledge work in the generative AI age. Your ability to generate high-quality prompts will determine your effectiveness in leveraging these powerful tools. Stay focused and Keep Building 
Chapters 00:00 OpenAI ChatGPT Pro 01:10 The Meta Prompt 09:18 Why Meta-prompting is important 11:36 o1 vs o1 Pro Mode - HN Post 18:53 o1 vs o1 Pro Mode - Content Gen 27:05 o1 vs o1 Pro Mode - AI Code Review 32:20 ChatGPT Pro - Worth? #promptengineering #llm #aiengineeringIf you want a more general approach to prompting ChatGPT when you're unsure of what specific elements or information you need to extract from an HTML DOM, you can use broader prompts that encourage ChatGPT to analyze the DOM and suggest relevant elements or patterns. Here are some general prompts you can use:
If you want a more general approach to prompting ChatGPT when you're unsure of what specific elements or information you need to extract from an HTML DOM, you can use broader prompts that encourage ChatGPT to analyze the DOM and suggest relevant elements or patterns. Here are some general prompts you can use:
- Explanation: This prompt asks ChatGPT to analyze the DOM and suggest potentially relevant elements without specifying what you need. It’s useful when you’re exploring the DOM for the first time.
- Explanation: This prompt asks ChatGPT to identify common patterns (e.g., user information, timestamps, images) that are often extracted from web pages.
- Explanation: This prompt encourages ChatGPT to act as a guide, helping you explore the DOM interactively and suggesting elements as you go.
- Explanation: This prompt focuses on identifying different types of data (e.g., text, images, links) that can be extracted, without requiring you to know specific element names.
- Explanation: This prompt is useful for identifying elements that are likely to change or contain user-generated or dynamic data.
- Explanation: This prompt asks ChatGPT to look at the semantic meaning of the DOM (e.g., headings, articles, metadata) and suggest elements that might be relevant.
- Explanation: This prompt turns the interaction into a dialogue, where ChatGPT asks you questions to better understand your needs and then suggests elements.
- Explanation: This prompt asks ChatGPT to suggest use cases (e.g., scraping user data, extracting images, analyzing timestamps) and then recommend elements based on those use cases.
- Explanation: This prompt focuses on identifying elements with unique attributes, which are often used to target specific data.
- Explanation: This prompt asks ChatGPT to analyze the DOM from the top down, suggesting elements at each level of the hierarchy.
Response:
"Based on the DOM structure, here are some potentially relevant elements and why they might be useful:
1. Tweet Text: Look for elements with attributes like `data-testid="tweetText"`. These likely contain the main content of the tweet.
2. User Information: Elements with `data-testid="User-Name"` or similar attributes often contain the user’s name and profile link.
3. Timestamps: Elements with `<time>` tags or `datetime` attributes are likely to contain the tweet’s publication date and time.
4. Images: `<img>` tags with specific `alt` attributes (e.g., `alt="Image"`) are likely to contain embedded images in the tweet.
5. Links: `<a>` tags with `href` attributes are likely to contain links to the tweet or user profiles.
These elements are commonly used in web scraping to extract structured data from social media platforms like Twitter."
- Exploratory: They encourage ChatGPT to analyze the DOM and suggest possibilities.
- Interactive: Some prompts allow for back-and-forth dialogue to refine your understanding.
By using these general prompts, you can effectively explore and extract relevant data from an HTML DOM without needing to know the specifics in advance.
General Prompts for Analyzing HTML DOM with ChatGPT
If you want a more general approach to prompting ChatGPT when you're unsure of what specific elements or information you need to extract from an HTML DOM, you can use broader prompts that encourage ChatGPT to analyze the DOM and suggest relevant elements or patterns. Here are some general prompts you can use:
1. Open-Ended Analysis Prompt
- Prompt: "Given this HTML DOM structure, can you analyze it and suggest what elements or data might be relevant to extract? Please explain why they might be useful."- Explanation: This prompt asks ChatGPT to analyze the DOM and suggest potentially relevant elements without specifying what you need. It’s useful when you’re exploring the DOM for the first time.
2. Common Patterns Prompt
- Prompt: "What are some common patterns or elements in this HTML DOM that are typically useful for data extraction? Can you list them and explain their purpose?"- Explanation: This prompt asks ChatGPT to identify common patterns (e.g., user information, timestamps, images) that are often extracted from web pages.
3. Interactive Exploration Prompt
- Prompt: "I have this HTML DOM, but I’m not sure what to extract. Can you guide me through exploring it step by step and suggest what might be useful?"- Explanation: This prompt encourages ChatGPT to act as a guide, helping you explore the DOM interactively and suggesting elements as you go.
4. Data Types Prompt
- Prompt: "What types of data (e.g., text, links, images, timestamps) can be extracted from this HTML DOM? Can you list them and provide examples of where they might be found?"- Explanation: This prompt focuses on identifying different types of data (e.g., text, images, links) that can be extracted, without requiring you to know specific element names.
5. Dynamic Content Prompt
- Prompt: "This HTML DOM seems to contain dynamic content. Can you identify elements that are likely to hold dynamic data (e.g., user-generated content, timestamps, or interactive elements)?"- Explanation: This prompt is useful for identifying elements that are likely to change or contain user-generated or dynamic data.
6. Semantic Analysis Prompt
- Prompt: "Can you analyze the semantic structure of this HTML DOM and suggest elements that might represent key pieces of information, such as user details, content, or metadata?"- Explanation: This prompt asks ChatGPT to look at the semantic meaning of the DOM (e.g., headings, articles, metadata) and suggest elements that might be relevant.
7. Interactive Query Prompt
- Prompt: "I’m not sure what to extract from this HTML DOM. Can you ask me questions to help narrow down what I’m looking for, and then suggest relevant elements?"- Explanation: This prompt turns the interaction into a dialogue, where ChatGPT asks you questions to better understand your needs and then suggests elements.
8. General Use Case Prompt
- Prompt: "What are some general use cases for extracting data from this HTML DOM? Based on those use cases, what elements should I focus on?"- Explanation: This prompt asks ChatGPT to suggest use cases (e.g., scraping user data, extracting images, analyzing timestamps) and then recommend elements based on those use cases.
9. Attribute-Based Extraction Prompt
- Prompt: "Can you identify elements in this HTML DOM that have unique attributes (e.g., `data-testid`, `class`, `id`) and suggest how they might be used for data extraction?"- Explanation: This prompt focuses on identifying elements with unique attributes, which are often used to target specific data.
10. Hierarchical Exploration Prompt
- Prompt: "Can you explore this HTML DOM hierarchically, starting from the top-level elements, and suggest what might be useful to extract at each level?"- Explanation: This prompt asks ChatGPT to analyze the DOM from the top down, suggesting elements at each level of the hierarchy.
Example Interaction:
Prompt: "Given this HTML DOM, can you analyze it and suggest what elements or data might be relevant to extract? Please explain why they might be useful."Response:
"Based on the DOM structure, here are some potentially relevant elements and why they might be useful:
1. Tweet Text: Look for elements with attributes like `data-testid="tweetText"`. These likely contain the main content of the tweet.
2. User Information: Elements with `data-testid="User-Name"` or similar attributes often contain the user’s name and profile link.
3. Timestamps: Elements with `<time>` tags or `datetime` attributes are likely to contain the tweet’s publication date and time.
4. Images: `<img>` tags with specific `alt` attributes (e.g., `alt="Image"`) are likely to contain embedded images in the tweet.
5. Links: `<a>` tags with `href` attributes are likely to contain links to the tweet or user profiles.
These elements are commonly used in web scraping to extract structured data from social media platforms like Twitter."
Why These Prompts Work:
- Flexibility: They don’t require you to know what you’re looking for upfront. - Exploratory: They encourage ChatGPT to analyze the DOM and suggest possibilities.
- Interactive: Some prompts allow for back-and-forth dialogue to refine your understanding.
By using these general prompts, you can effectively explore and extract relevant data from an HTML DOM without needing to know the specifics in advance.
Final Thoughts:
These prompts are designed to help you navigate the complexities of HTML DOM structures, especially when you're unsure of what to extract. By leveraging ChatGPT's ability to analyze and suggest, you can save time and focus on the most relevant data for your needs.Happy Scraping!