Dave Shumaker: Rochelle, to start our conversation, could you briefly recap the themes you spoke about in the Beyond Publishing webinar?

Rochelle Grayson: Sure. There’s a lot of fearmongering about AI, especially in creative industries like art and publishing. Chris reached out to me because I had done a previous webinar about the impact of AI tools on creative artists and writers. But I also wanted to speak to the impact on the business of publishing. The use of AI isn’t just about the creative products; it’s also in the internal processes and functions that can free up time for being strategic and building the business. My goal was to broaden the discussion beyond the use of AI in the creative process. An example would be the use of AI in data analytics to inform marketing.

Shumaker: That’s a great lead-in to my next question. You co-founded an organization called Circles of AI. What’s the scope of that organization, and what does it provide?

Grayson: I started Circles of AI with a colleague after I read a study that analyzed the adoption of AI. One of its many findings was that women were far behind their male counterparts in leveraging AI, even in the same job. I was shocked that a gap was already forming. I was also seeing that large corporations were outpacing small and medium-sized businesses in leveraging AI. I felt that was a missed opportunity for small businesses, because they could potentially get the most benefit from these tools. The goal of founding Circles was of AI to provide an entry point for the people who were being left behind. It serves people coming from a small business or nontechnical perspective. It can level the playing field, help the people being left behind to innovate, and even leapfrog the leaders.

Shumaker: How long has it been in operation, and what’s your experience been like so far?

Grayson: We launched in June, so it’s still very new. We have about 150 members from a wide range of industries, including a few from publishing. We have a WhatsApp group where people have conversations and ask questions. It’s both casual and educational. We’ve also been doing weekly webinars where we tackle aspects of AI as they relate to business processes and workflows. We talk about specific tools, but we don’t try to be exhaustive.

Shumaker: Is the membership mostly in Vancouver, or is it widespread geographically? What are your plans to spread and grow it?

Grayson: Actually, just a small percentage are based in Vancouver. Our growth has come from contacts and colleagues from all over. We have quite a few people from the New York region and around the East Coast. As for promoting growth, we’re planning some initiatives for the fall.

Shumaker: With the emphasis on peer-to-peer communication and learning, have you seen clusters forming around specific industries or functional interests?

Grayson: So far, we’ve seen some clusters starting to form, but the group is still small enough that it’s manageable as one group. We want clusters to form organically—that’s why we called it “Circles” of AI. We’ll still keep a general circle for the basics. Within affinity circles, members can explore in depth. There, the emphasis will be even more on peer-to-peer learning. We have a tagline—“From learning to leading together”—that captures the idea that one day you might be learning from someone else, and the next day leading with your own expertise.

Shumaker: Have you been able to judge how far along the adoption curve your incoming members are? Are they at the stage of curiosity and considering how they might benefit from AI, or have some already gained experience with adopting and using it?

Grayson: Many members are very early on the adoption curve when they join. Maybe they’ve used ChatGPT a couple of times and didn’t get very good results. They don’t understand what all the hype is about, because AI didn’t help them. They have a mixture of curiosity and skepticism. We show them a variety of ways they might benefit from it—a variety of areas where they could implement it. We recommend that they start by picking one area to experiment with. Many times, it’s an area where they can see measurable productivity or financial gains. With my own financial background, I want to make sure that the tool is actually helping move the needle on the business. An example is doing financial projections. We don’t recommend trying to apply AI to multiple functions all at once.

Shumaker: Rochelle, as I learned about your history, I noted that you have a background in ethics and have taught ethics at the university level. Are ethical concerns on the radar of new members of Circles of AI, or is ethics something they become sensitized to as they gain experience?

Grayson: There’s been so much talk about ethics and bias that the concerns keep some people from trying to use AI. In our community, we try for a balance. As we talk about AI applications, we bring up the benefits and also the risks, including the ethical concerns. Some of the considerations that come up are disclosure (making it clear how and when you are using AI) and privacy (the limits on sharing personally identifiable information). We want our members to be good AI citizens who are accountable for privacy and security.

Shumaker: In addition to Circles of AI, I believe you’ve been doing some work with young people. How is their perception of AI?

Grayson: Yes, I taught at a summer camp, where I had 45 high school students who were 15–17 years old. A lot are excited by the potential of AI. They’re also at a point in time when everything is at a crossroads. Their teachers are nervous because they think students will use AI to cheat, so it’s been pounded into them that if you’re using AI, you’re cheating. My approach was to teach them how to use AI responsibly—as a great creative assistant, but not to copy and paste AI output. Instead, read it, validate it, and review it critically. AI requires new skills: critiquing, evaluating, and problem-solving. What to ask—inquiring, the skill in framing questions, or “prompt engineering”—is also a new skill for many students. So, the education system will need to teach these new skills so that students leverage the new AI technology.

Shumaker: As we close, is there anything else you’d like to add?

Grayson: I feel we need to address the bias prevalent in many AI models. People complain that the models are biased, but they’re only highlighting the bias that’s prevalent in society. They’re reflecting us. So, we need more content that provides alternative perspectives, and we need to train the models on that. Also, we need people who constantly bring up the issue. If we are to change things, everybody needs to have a voice. Who’s making the decisions? If the powers that be continue to be the same powers that be, then the models will reflect their view of the world. Our greatest threat is that people believe computers because they perceive them as objective, but in fact, their results are based on the opinions and perspectives of those who train them.