Yes. Why? From an Insider article earlier in the year, ‘ChatGPT could be used for good, but like many other AI models, it’s rife with racist and discriminatory bias‘: - Large, uncurated datasets scraped from the internet are full of biased data that then informs the models.
- ChatGPT itself is trained on 300 billion words, or 570 GB, of data.
- Like many AI models before it, bias can be found in its output.
- Researchers warn its real-world applications could spell trouble for marginalized groups.
Read more here. | |
Report bias in responses to ChatGPT prompts. Email ethics@openai.com |
| | Worth a watch: The video below (23 minutes long) asks ‘Are We Automating Racism?‘ | |
Addressing AI Bias in the Classroom | |
🤖 Talk about it with students. 🤖 Give students tasks such as “Spot the Misinformation” to highlight bias in AI. 🤖 Students need to know how Large Language Models, such as ChatGPT, work. ChatGPT can help you explain it to them – see below! | |
🤖 Better still, highlight the bias. ChatGPT can help you explain that to them too! | |
The European Commission has published ‘Ethical guidelines on the use of artificial intelligence (AI) and data in teaching and learning for educators‘. Available in 24 European Languages, including Romanian, here. | |
If you are keen to learn more about everything AI, then this is the Padlet for you! It was curated by Dr. Ashley Cross, Senior Director of Edu at ATLIS. Ready for a rabbit hole?! | |
MORE PROFESSIONAL LEARNING | |
Write Better ChatGPT Prompts | |
Check out this Twitter thread from Nicolas Cole. In his words, “The best AI writing prompts aren’t 1 sentence. That’s like having a Ferrari and driving it 5 mph. Here’s a 60-second masterclass on how to create prompts like a virtuoso writer.” | |
|
|