Learn • Anywhere
Rumie logo

Are you up for an incredible Artificial intelligence (AI) adventure?

Stan Marsh from South Park says 'ChatGPT, Dude'  to a confused peer.

Say hello to ChatGPT, one of the largest language models and your virtual companion. ChatGPT can process and generate text quickly and well, but...is ChatGPT safe?

Safety is key on this digital journey.

So let's buckle up and figure out how to use ChatGPT safely and effectively.

All aboard!

Did you know?

"A language model is a machine learning model that predicts the next word given a sequence of words."

1. Verify all information

ChatGPT is a language model, NOT a source of truth.

Trust No One Swipe Left GIF

ChatGPT is a powerful tool, but it learns from the data we feed to it.

This means that it might generate answers that are not accurate.

Do your own research and confirm all information.

For example, if you ask ChatGPT what is the tallest building in the world, it will tell you that it's Burj Khalifa in Dubai, United Arab Emirates. To confirm this, you can:

  • Google the building's name

  • Ask an architecture professor or another expert

  • Ask ChatGPT to cite its source and check the source directly

2. Don't overshare

Never share personal or confidential information with ChatGPT.

Tom Cruise in Top Gun refuses to answer a question saying 'It's classified.'

How can you tell if you crossed the line?

If you wouldn't print it and leave it at your local bus stop, then you shouldn't feed it into ChatGPT.

Quiz

What information is OK to enter into ChatGPT?

Did you know?

In May 2023, Apple joined the growing list of companies restricting the use of ChatGPT for work in order to protect their proprietary information.

Queries we enter into ChatGPT are visible to OpenAI (the service provider), stored, and will probably be used to develop the service at some point.

3. Be clear and concise

The clearer and more detailed your instructions, the more relevant the response you get.

Okay. Maybe I am not expressing myself clearly.

ChatGPT can't think for itself and it can't read your mind. You need to tell it what you want it to do.

If you aren't clear, you may get wrong or irrelevant results.

If your topic is too complex, break it down into smaller parts to improve your results.

Quiz

Anya wants to learn about World War II. She asked ChatGPT for help: "Tell me about the history of the world." This gave her an outline of human history from ancient to modern times. How could she improve the prompt? Select all that apply:

4. Promote ethical AI use

AI models like ChatGPT can generate realistic text, including fake news, misinformation, and biased content.

Hands with controller. Text says 'Save the world from misinformation; pause before you share.'

What should you do?

  • Verify information before sharing it

  • Reportharmful outputs to OpenAI

  • Educate others about the limits and responsible use of ChatGPT

Did you know?

To evaluate content, look to the right of the generated text and click the thumbs up or thumbs down icon. Any details you share will help improve future versions of ChatGPT.

Take Action

Chatbots are only as smart as the people who build them

Correction: Chatbots are only as smart as the people who USE them, so go be a responsible ChatGPT user!

License:

This Byte has been authored by

AD

Anita Damjanovic

Customer Education Specialist | Educator

PhD

English

🍪 We use technical and analytics cookies to ensure that we give you the best experience on our website. more info