I love that AI can see or find patterns. The advancements that can mean for everything from space exploration to medical treatment to analysis of ancient art are glorious to think about. I also love that AI can apply grammar and spelling rules to things I’ve written.
This is from the 2025 Human Development Report from the United Nations Development Programme:
AI-powered identification of bird species can enable users to better identify them in the future. The increased ability to identify species can, in turn, improve AI’s performance by directly contributing geolocated observations and uploading labelled sounds and images.
I love all of this.
However…
Have you ever seen a movie that you were assured by a trusted film critic was horrible and, in fact, it turned out you loved it? Or the opposite: he said it was wonderful and you loathed it?
Have you ever been told by a few people that someone you haven’t met yet through work or in your neighborhood was boring and not really worth hanging out with, but you did finally meet them and you thought they were hilarious and fun to be with?
Have you ever read a summary of a long report or a book, but then you read the report and you came to a very different conclusion than the summary?
Those scenarios are why I don’t trust AI for decision-making or for all things related to customer service and most client interactions. Those scenarios are why I want to keep interacting with clients and the community myself.
Whether I’m the director of a project or directing communications for a project, one of the first things I like to do is take some time to answer questions that come in via phone or email. It never turns out quite like it’s been portrayed to me, especially by an over-worked, under-trained administrative assistant or an executive director who has a million other priorities. Nothing teaches me more quickly what’s been working and what has not. Sure, I might come up with canned responses to frequently asked questions, but if that response doesn’t answer the person’s inquiry, I want to know about it.
And then there is the idea of care and support: for clients, for volunteers, for the community. Care is about awareness, concern and looking after something. Providing clear, easy-to-access and easy-to understand information for an audience is a part of caring about that audience, but it’s not all of it, because quality, sincere care – the kind that keeps clients coming back, that keeps volunteers engaged and keeps donors supporting the organization – is emotional and relational. When I listen to certain staff members talk on the phone with the clients of one of the nonprofits where I work, I know that AI is no substitute for those interactions – and that if we stuck those folks on an automated phone system or chatbot, we’d not only let them down, we would lose them.
I also would never trust AI to author most social media posts, nor answer most messages that come in via social media or an online group: I would never cede the incredible knowledge that comes from interacting with people, with actual humans, to AI.
I can’t imagine taking away one of the best ways ever to gather knowledge and build skills: interacting with people MYSELF.
I wrote about this back in 2017, in fact: No app can substitute for actually talking with people. The Tech Bros must really loathe talking to actual humans.
Also see:
Artificial Intelligence – friend or foe for nonprofits?
schedule social media posts? use with caution
If you have benefited from this blog or other parts of my web site and would like to support the time that went into researching information, developing material, preparing articles, updating pages, etc. (I receive no funding for this work), here is how you can help.