top of page
Search
Writer's pictureLindsay Morgia

99 Problems and AI is one (Part II): Proceeding with caution

Updated: Aug 12

A few weeks ago, I attended the Nonprofit Technology Conference in Portland, OR. It was my first time attending, and it was a wonderful experience. Interestingly, almost every session I attended discussed the pros and cons of using AI, even if the session wasn't about AI at all. So, did the experience ease my concerns about the technology? No, not really, and here's why.


A red fingerprint that appears to be emerging from a clipboard. Image AI-generated.
An AI-generated image for "Data privacy."

Top problems with AI

First, in an earlier post about being an AI skeptic, I mentioned that one of my biggest concerns about using AI for client work is data privacy (or lack thereof). The conference sessions I attended confirmed my concerns, highlighting the privacy and security issues surrounding AI technology. Also, since government regulations have not caught up to AI technology, keeping track of different privacy policies (if any) that protect people's personal information is complicated. That may start to change thanks to the European Union's AI Act, but whether the United States will follow suit is debatable.


Second, several presenters discussed using AI as a "thought partner." Sure, we can ask AI about initial trends in our data or feedback on how we should present specific findings. But, as the team at the CT Data Collaborative pointed out, AI does not have the historical, social, cultural, or political context for the data-related questions we ask it. As such, what AI thinks is important may be completely irrelevant and could lead organizations down the wrong path in finding solutions.


A robotic lion with a headset, wires holding up the next, and a breastplate that says "JCHVI." AI-generated image.
An AI-generated image for "Be nice to AI." Terrifying!

Finally, do I think AI will eventually become our robot overlords? I'm not sure, but a conference participant mentioned a Microsoft report that says if you are nice to your AI tool, it will be nice to you, so…yikes?  But despite my preference for human interaction, AI is becoming increasingly unavoidable. So, in the spirit of transparency, here are the ways I am cautiously using AI with clients and in my business.


AI for data analysis

When I work with nonprofits to conduct interviews and focus groups, I use two tools that have AI features:

  • Otter.ai is a program that transcribes interview and focus group recordings into text

  • MaxQDA is a qualitative analysis software that helps analysts identify themes across conversations

You may have seen OtterPilot pop up in Zoom meetings, which captures conversations and summarizes the key points. Otter.ai uses AI to do the same thing for interview and focus group recordings, but I still find it necessary to review the summaries and transcripts to catch errors. The newest version of MaxQDA also has optional AI features that I use on a case-by-case basis, like the ability to summarize long passages of text into key points. However, I always double-check to ensure the AI suggestions make sense.


I feel more comfortable using these tools because their privacy policies follow the EU's General Data Protection Regulation standards (GDPR). The GDPR is Europe's data protection and privacy policy, which is generally considered the gold standard globally.  


AI for report writing

I have been a fan of Grammarly for years.  As someone who works by myself and does a ton of writing, Grammarly has been so helpful in catching mistakes and telling me when I'm taking too damn long to get to the point (shoutout to all the other recovering academics!). Grammarly is HIPAA compliant, FERPA compliant, and complies with GDPR.[1] The program isn't perfect, and its AI features have made some strange suggestions over the years. Recently, I have been considering hiring an editor to discuss improving my writing instead of relying on the software to do it for me.


AI for marketing

I have zero background in marketing, so I'm always looking for guidance on creating social media posts and website content. I have used ChatGPT for social media suggestions, but I don't think I've ever used the results word for word. I certainly see the value of having a tool that can help generate marketing ideas. However, I don't know that I love the suggestions enough to use ChatGPT regularly. Besides, I enjoy being a little loopy on social media, as I believe in authenticity. 😊


Where do you stand on the use of AI in your organization? How has it made your work easier, and what problems worry you the most about using AI? Leave a comment below to share your thoughts and experiences!


[1] HIPAA stands for Health Information Portability and Accountability Act, which protects patients’ healthcare information. FERPA stands for the Federal Education Rights and Privacy Act.


15 views0 comments

Comentários


bottom of page