Using AI with Ethics in Mind
AI is everywhere. We are reaching a point where, whatever personal misgivings we might have, using AI remains an essential skill. There is no denying the power of generative AI. But it is also clear that AI has issues. There are real ethical concerns at play: environmental impacts, data privacy and security, inaccuracies in generated content, and perpetuating bias, to name just a few. So, how do we find a balance? How do we use AI ethically?
Here are a few ideas to help you use AI while keeping ethics in mind.
- Take time to reflect. Before using AI for any task, think about the impact. Ask yourself how using AI for this particular task will benefit different stakeholders (e.g., instructional designers, learners, instructors, administrators, etc.) and how it might harm them. Weigh whether the benefits outweigh the potential for harm. You should also take a moment to consider whether AI is the best tool for this particular task before proceeding. Just because you can use AI does not necessarily mean it is the best option or that you should use it.
Tip: Check out an article about what AI can and can’t do through the lens of Bloom’s Taxonomy: ChatGPT and Bloom’s Taxonomy: What AI Can and Can’t Do. - Be concise and minimize prompts. Every word and every prompt you type into ChatGPT or other AI platforms has an energy cost. The servers that process them use electricity, and they use significant amounts of water for cooling. Images have even larger energy costs. Aim to minimize the length of your prompts and the number of prompts you need to get your desired output. Use generated images sparingly and intentionally.
Tip: If it takes you several prompts to get to a desired output, consider asking ChatGPT to generate a single, concise prompt that would get a similar output for you to use next time. - Audit your outputs. AI hallucinates. It makes stuff up, and it’s really good at confusing correlation with causation. It’s also only as good as the data it was trained on, and that data, unfortunately, is often biased. When generating content, especially content used in educational contexts, it is really important to verify and fact-check. Make sure generated content is reviewed for both accuracy and bias.
Tip: Check out an article about how to minimize hallucinations with AI: Taming AI Hallucinations: A Guide for Instructional Designers. - Be transparent. Include a statement about the use of AI as part of your course design. Let people know if and when AI is used to generate content, and how AI is incorporated elsewhere. Ensure everyone is given an opportunity for informed consent if any of their data will be used by AI.
At the end of the day, I aim to use AI impactfully and intentionally. How do you approach using AI with ethics in mind? I’d love to hear your ideas.
Want to Learn More?
Connect with your institute's D2L Customer Success Manager or Client Sales Executive, or reach out to the D2L Sales Team for more information about how Learning Services can support you on your learning journey.
Comentários
-
Great tips, Andrea! To point 2, I'm learning that as AI gets more efficient, its environmental cost has also decreased. According to AI researcher Ethan Mollick, "Google has reported that energy efficiency per prompt has improved by 33x in the last year." He also notes that a standard prompt "is roughly 0.0003 kWh, the same energy use as 8-10 seconds of streaming Netflix or the equivalent of a Google search in 2008 (interestingly, image creation seems to use a similar amount of energy as a text prompt)."
Anyway, thought you might be interested. Certainly still has an energy cost but at least it's getting better.
