Everyday AI: What awesome tricks has ChatGPT mastered?

ChatGPT can do more than answer questions and compose letters or emails. Robots have other abilities, sometimes surprising and useful in everyday life. By the way, some tricks also work with Microsoft’s Copilo AI app and Google Bard – it’s worth a try!

Present information in an attractive manner

ChatGPT converts existing information into a visually appealing form: on request, the bot converts the list into a paragraph of written text. Or it formats nicely as a table. For this, the AI ​​can identify how the columns should be distributed. If it doesn’t work the first time, give ChatGPT some guidance on how to fix those errors.

For example, you can tell it that you want to get the answers in table form. This also works with complex tasks like: “I want you to create a table of all major cities in Switzerland, with columns for the number of inhabitants and the year they exceeded the 100,000 threshold inhabitants.”

These tables can be selected, copied and processed in Word, Excel, Google Docs or other Office applications. They are child’s play for bots and can be reused in office documents without any problems.

Assists with effective computer use

The bot is also proficient in computer and programming languages. If you have a website to manage, ChatGPT can provide tables in HTML code. In a spreadsheet, you can get help with formulas.

ChatGPT can also help you with programming languages. You may think that this does not benefit you as a non-programmer. But this is not the case, quite the opposite. Thanks to AI, even computer users without specialized knowledge can use the programming functions of Mac, Windows and Office to automate common tasks.

Here are two examples:

Work with images

AI now processes not only text, but also images. Microsoft’s Copilot app allows you to upload photos just like ChatGPT 4, which requires a paid subscription.

After uploading the image, ask questions such as: “What is the architectural style of the row of houses on the right?” ChatGPT creates image descriptions, captions and provides the keywords needed to integrate photos into the image bank.

Or take a photo as input for computer-generated graphics. Transform a selfie into a cartoon character or instruct ChatGPT to artificially age or rejuvenate the person depicted. For landscape photos, you can change the weather or mood. Also fun: upload a photo of where you live and ask, “What will this place look like in a hundred years?”

The results of these types of experiments are often uncertain: AI does not seem to process images directly, but through textual descriptions, which can reduce the recognition value. Nevertheless, this is an exciting area of ​​experimentation.

Check through the documents

In addition to images, the bot also accepts PDF files. This requires GPT 4. If you don’t have a paid subscription, try turning on the PDF feature confusion.ai. After uploading a file, you can summarize it and ask targeted questions about its content. As always with AI, trust the answers only after checking them – in my tests it happened early on that the AI ​​not only refers to the content of the document, but also consults other sources. In the case of use as a homework aid, this would be fatal.

Robots for special tasks

Finally, ChatGPT can be “trained”. He is trained for specific tasks. The possibilities offered are visible in the “Exploration” section in case of paid subscription: there are serious experiences and others that are more entertaining.

Coloring book heroes are more in the latter category. Based on the description you give it creates colorful pictures for your kids.

For someone like me, who doesn’t have a green thumb, Planty can prove to be important. This bot provides information about plant care.

Did you get an error? Please inform us about it.

5 comments