ChatGPT can now see, hear, and speak

In my nano tips series on ChatGPT so far, I’ve covered Data Storytelling and Visualization, and Technical Prompts. Since “ChatGPT can now see, hear, and speak”, I thought it would be a good idea to create some additional lessons to explain those multimodal capabilities. I've included some of my favorite examples on prompts as well.

The full list of tips and tricks can be found in my LinkedIn Learning course called "Nano Tips for Leveraging Technical Prompts Using ChatGPT with Lachezar Arabadzhiev".

1. Join Multiple Tables With Prompts

Easily merge two data tables by using ChatGPT. Simply upload the tables and instruct ChatGPT to join them using the ID column as a unique identifier. ChatGPT will analyze and merge the tables efficiently, providing you with a combined file for download. This method eliminates the need for specific coding or queries, making complex data operations accessible to everyone, including those without technical expertise.

2. Analyze Data With Linear Regression

Harness the power of ChatGPT for advanced data analysis like linear regression and correlation studies. Begin by uploading your dataset and clearly stating your query. For instance, explore the relationship between experience and salary in your dataset. ChatGPT can quickly perform linear regression and other analytical techniques, delivering prompt results.

3. Extract and Organize Text From Images

Leverage GPT-4's intelligence combined with ChatGPT's Vision to interpret images effectively. For confusing parking signs, simply upload a photo to ChatGPT for clear parking instructions, avoiding unnecessary fines. Additionally, ChatGPT can digitize handwritten content. Upload a photo of a handwritten table, and ChatGPT will transform it into a digital format, even offering a CSV file option.

4. Code an App Element From an Image

Utilize ChatGPT's vision feature to convert images into code for web elements. By uploading an image of a desired element, like a button, into the ChatGPT mobile app, you can request the corresponding CSS, HTML, or other relevant code. The code generates instantly, which you can then test in tools like JSFiddle to see a similar working element. While not always a perfect match, it serves as an excellent starting point for web development, applicable to various elements and potentially entire pages.

5. Voice Conversations and Commands

Now let's talk about the use of voice in the ChatGPT universe. To get started, open the mobile app, then tap the microphone icon and begin speaking. The app will process your voice prompt in real time and transcribe it into text using Whisper, OpenAI's open-source speech recognition system.

Thank you for making it this far and I hope that with each issue of the "My Next Story Is..." newsletter, you would be able to learn something useful from my experiences. Here are a few additional resources to help you further your learning.

? Explore my LinkedIn Learning courses on storytelling through data and design!

? Follow me on LinkedIn,?and click the ?? at the top of my profile page to stay up to date with my latest content!

? Subscribe to receive the by-weekly "My Next Story Is..." newsletter.