This week on EthicalVoices, Loren King, digital marketing supervisor for AI and insights at MorganMyers, discussed ethical issues including:
- How coffee talks can build a healthy AI adoption culture
- Why organizations need an AI philosophy, not just policies
- Why communicators must balance AI innovation with humanity, empathy, privacy, and environmental impact
Tell us more about yourself and your career.
I fell into AI from a place that you wouldn’t expect. Morgan Myers is an agency that focuses on food and agriculture clients. We have around 40 years of history working with everybody from Merck Animal Health Cattle to commodity associations like Iowa Corn. Essentially it’s those who support farmers or put food on the table -like Bob Evans and post-consumer brands.
I come from a farming background. My family raises corn and soybeans in southwest Michigan. So for me, it’s a little bit of a change around to end up talking artificial intelligence all day. But I’ve always been naturally technology curious and philosophy and ethics curious, which is why I’m so excited to be here.
I paid attention to what was going on. I did some independent learning. In my degree program, which actually focused on education, I had some advisors who supported going into more technology spaces. So, as AI became talked about and when it came to researchers and Google was releasing things around identifying cats in early 2015, I was watching what was going on.
I joined MorganMyers, and six months later, ChatGPT dropped into the world. We needed a point person, and I had the knowledge to work on it, which has evolved into running dedicated projects, supporting clients on understanding, and just generally trying to keep up with the news.
People don’t realize just how tech savvy farmers are. My father has sold crop insurance in Iowa. Farmers are actually early adopters in a lot of cases.
We just finished up some meetings this week. I was out in DC with a technology advisory group for the American Farm Bureau, and the people on that group are farmers or run agribusinesses, and many of them have incorporated AI directly either into monitoring inventory or speeding up the receipt processing or even asking questions about the crops in the field as they’re standing there. There’s many use cases, and it’s fun to watch that group of early adopters really shine.
What is the most difficult ethical challenge you ever confronted at work?
This is something that has emerged in conjunction with AI, and it’s keeping ethics top of mind at all times. When we started planning for a real AI integration at Morgan Myers and then at GS IMCG, which is our group, and at GSBC, which is our sister agency under that label. There’s two agencies working together in one business, and as those conversations kicked off, it was right in our face.
Should we prioritize efficiency at all costs, like many businesses do? How do we balance the creativity and the expertise that we bring and augment the capacity of our employees against some of the real gains that we can have.
It was building out frameworks and training that muscle of considering ethics with really every single decision we made. Drafting AI policies. Speaking to a client. I’ve had client interactions where I’ve done a presentation and afterwards they come to me and they say, what about X item?
And it’s always impressive because I’m expecting, what’s the best AI tool for me to use? Or what’s one thing I can teach my employees? Actually it’s usually a question around philosophy or ethics or what is the right way to approach this. And so that muscle has developed over time for making a continual ethical choice that not only our AI enablement team which I get the chance to sit on is thinking about, but also our client leads and or our individuals who are using these items. I’ll give a good example of how this works out.
I’ve been trying to integrate AI into helping draw insights as we’re working on reports. But I don’t want to lose my ability to come up with those conclusions myself. So I started building rules of thumb. One out of every three bullet points for insights I want to have either generated by an AI tool or at least having an AI tool come up with something that reinforces an idea I already had.
At the same time, if we’re drafting content, we want to have a rule of thumb for three months of human written content to train an agent before we use it for anything, because helps with authenticity, but it also keeps the muscles of our writers strong when it comes to drafting something.
Those are just a couple examples where we keep in mind ethical frameworks. AI impacts humanity more than just the next great technology tool. Yes, an iPhone is awesome, but an iPhone can’t build itself or rebuild itself in the way that an AI tool potentially could.
I’m a firm believer in keeping ethics top of mind. How do you help your organization keep it top of mind and work with staff help them know what questions to ask and then remember what they should be asking?
We found it really helpful to identify those folks who have a natural interest in AI or advancement, or are just a curious. We then pull them in and make them dedicated AI champions.
This doesn’t mean they know everything or have extra resources, It just means that they’re willing to talk through. Using AI or not using AI.
The second part of this is building out guard rails and letting people know, the positives and negatives that might emerge from the use of AI.
It’s just being realistic with people on what the concerns are. For us, the most effective way to do so is coffee chats. These coffee chats happen once a month. They have a very basic framework within the coffee chat for what we’re going to talk about each time so everybody knows what to expect. We kick it off with a little bit of news, just what’s the big thing you need to know? ChatGPT 5 has launched, or 5.2 Claude bots are out in the world.
Stuff that might be applicable for others. Then we will have an employee highlight a way they’re using AI. It can be super basic. One of our favorites is always just summarize my week. It’s a pretty helpful one where people will talk about how they use it to go through their emails and their SharePoint files and remember what they did for the week so they can pull that up and just finish their day or their Friday evening in the right way.
Then we’ll usually focus on an area that’s applicable for either a client or a sector. For example, E-commerce and AI and how that’s going to impact food clients. What are the ways it’s being used on the farm? How is AI impacting influencers?
This is very intentionally low key. There are guardrails, but there are not KPIs . Anybody can come in and share their opinion. The chat’s usually pretty active.
There’s no requirement that you have to be there, but we get very strong attendance. As a result of that, people share ideas. These ideas and suggestions go to our more formal team, who can then say, yes, here’s the good consideration from that, here’s another way you could approach it or here’s a concern you might want to keep in mind for this when you’re presenting it to a client. It really has changed the game for us.
What I really like about it is you’re having non-AI champions lead it and share. We do something similar, but with different teams and practice areas because we know that the creative team has some very specific concerns so you can take a deeper dive on relevant issues.
We’re starting to encourage a little bit more of that even on the client level too. Some agriculture clients are going to be more conservative, about technology adoption. That’s just the cyclical nature of the business.
We might push them on some automations, but past that, we’re going to say, Hey, it’s all right. They’re not really ready to have their own chatbot on their website or they’re not interested in it.
Other clients are very innovative, and they’re coming to us with ideas and asking, Hey can you do this? And we’re saying. Probably. But have you considered doing a training on how you should think about AI before you’re actually deploying the AI project? So that’s the balance, and it does help us to have some staff who are trained up in AI tools. We can provide a lot of impact just in the same way that an SEO specialist would, except sometimes our recommendations cross over into the human and the education side of things more than you’d expect.
What are you seeing as some of the key ethics challenges for today and tomorrow?
Just keeping up with everything that’s going on. If somebody read about ChatGPT from 2022, they’d think, this is interesting. Maybe it can help me speed up my email writing, of course, as many people do.
Then if they read about when Cowork released and it could do functionally, anything a computer can do, how do you process that rapidity of change?
I’ve curated lists of newsletters and I follow all these people on social and I listen to the podcast and there’s still things that slip through the cracks for me that are huge, that are very important.
One strategy is to stay specific to your niche or your industry? There are not going to be as many immediate changes in something like chemistry or healthcare as there is just in AI in general.
Past that, you need to have a philosophy of how you want to approach it.
We’ve been drafting a group-level philosophy that says, this is how we approach AI. There are things we discard. We don’t necessarily consider X solution, and we’re not going to consider X approach because it doesn’t follow our philosophy.
It’s the philosophy that hopefully anybody can grasp and hold onto. It’s only a couple sentences. It’s nothing complicated. Instead of creating this rubric, the philosophy has started to act as like a north star on some of this.
Another challenge is the balance between generations. We have people from all generations using AI tools, but the way they think about them is different.
AI natives, Gen Z, Gen Alpha, kinda like myself. We’re used to engaging with a chat bot. We’re used to communicating natural language in that way. Others are used to maybe using more formal descriptions. Bridging that gap has been something that’s been on my mind a lot.
Cultural differences play a huge role as well. Your AI tool is personalizing to you as an individual, then it’s personalizing to somebody who’s a millennial or who’s Gen X, who likes media from those places who likes different media sources.
I can go on and on about the different challenges, but those are some of the ones that I’ve been thinking about the most recently.
The one I really like, is just dealing with the rapidity of change, the need to revisit what AI can do. But also always ask where does the data go? How do you keep it secure?
That’s where a strong AI policy can come into play. Philosophy can only go so far when it comes to actually understanding data sharing rules. That came up a lot in my conversations recently with people across the industry, what are the rules for having a research paper summarized?
It’s fantastic to take something that was written for a PhD level intellect and make it accessible for somebody like me on a topic along the lines of soybean cis nematodes, which we need to draft social for. But if that research paper isn’t published yet and we’re just preparing for the publishing and the announcements that’ll come with it.
We really don’t want it to be fully incorporated into ChatGPT’s models for anybody else that’s asking that question up until the formal release is available. That’s where we really rely on our IT team and some of our folks who can go in depth and say, look, here’s the proof of the privacy standard here,
One of the bigger challenges is there’s a lot of concern for folks about just the energy uses. Do you hear that from your team? Working with farmers, I could see that being a very relevant issue.
Oh yes. It’s actually probably the second or third most common challenge that’s brought up, whether it’s clients or internal teams or anyone. Water use is probably the biggest. Energy is also very important. The loss of farmland for something like a data centers.
I was talking to somebody this week who said they were feeling that there’d been an increase in hailstorms in the area since a very massive data center had been put in to support AI and was acting as essentially some type of heat sink.
I’m not here to speak to whether that is actually accurate or not, just to say that it is a legitimate concern that is being brought up and I think the answer is, we’re not putting this back in the bottle, but we can decide the lanes that it goes down, and so there are lots of situations where you can encourage less energy and water use.
It can be hosting on device. Much of my general day-to-day searching could probably be done with a few year old LLM that’ll get me the answers that I need and that doesn’t really use that much power. Apple is working on this partially with Apple Intelligence, caching responses.
I learned that Netflix does this where if you’re calling out to watch a show, there might be basically a storage device somewhere with the most popular shows in your community. And so instead of having to stream from somewhere else, use more energy up, it’s just available faster. The same thing is starting to happen with AI responses.
Some of the things that are searched the most, they’re just saving the response somewhere and then giving it back to you versus directly querying the LLM. So if I’m looking up burrito recipes or what’s this Nike shoe description? Those things are potentially not using as much energy as they would’ve in the past because it doesn’t necessarily have to query every single time.
Now, there’s the balance there then is, are you getting the most up-to-date information. It’s a pretty complex situation. We’re starting to think of AI as an infrastructure more than anything, and infrastructure has a different set of rules and a different history, a different impact on communities. Agriculture can speak to that with electricity and roads and dams and everything else. So I try to navigate it in the way of saying, yes, there is an impact. That doesn’t mean there aren’t solutions and innovation can at least lessen this impact.
The encouragement I have for some people is, look, if you don’t have to Google it or if you don’t have to look something up, don’t. Rely maybe on traditional search tools for something that’s basic and save your more complex queries for AI.
What is the best piece of ethics advice you ever received?
So this is a fun one . I came across it online – “All who are truly strong are kind.” It’s from Vagabond, but the idea has been expressed by many people in many ways, basically for the history of humanity.
And the reason that truly stands out to me is that it’s making you consider two things at once. This is what I think is distinct about humans. Everybody asks what makes us distinct from animals? Is it having opposable thumbs that we can really use? Is it the fact that we can plan?
And for me it’s that we can do two things at once. We can consider two items at once. And so when I’m approaching an AI project, I want to consider the absolute benefits of it, the strength. But I need to keep in mind the humanity, the kindness.
So that piece of ethical advice applies not just to AI, but to my personal life.
While I’m not perfect at it, I really do think that keeping that simple phrase in mind, and the fact that it considers two things at once is valuable in any ethical discussion.
I like the highlight on kindness, because kindness is something that can’t be faked.
Absolutely. Kindness is hard. It means usually making more time available. You need to be available for calls to just talk to people and to assuage concerns that maybe you don’t think are concerns, but others have.
And you have to think from their perspective. Empathy plays a role here, but kindness is like empathy expressed as an absolute in terms of the actual actions that you’re taking.
Is there anything I didn’t ask you that you wanted to highlight?
One thing we didn’t talk about too much is just the positive outcomes that can come out of all of this. Like the fact that we are considering ethics at all when it comes to a new technology like this.
If you compare to social media or even the internet, I am guessing the ethical discussions popped up way, way earlier when it came to AI. I’m really glad that people think in that way. I’m really glad that when I go to a client meeting, they ask that question and they say, not just, how is this going to impact my employees? But also how is this going to impact my mother who, lives alone? And maybe she could use this tool to do something good. Is it helpful for her? But maybe she shouldn’t, should she build that relationship with an AI tool or not?
There’s always discussion around what is a liberal arts degree for, what is a writing degree for, what is literature understanding for? Even from a technical perspective, it makes you better at using AI, but from an ethical perspective, it makes you better at knowing what AI can be used for and what it shouldn’t be used for.
The reemergence of the humanities is my favorite. Even if it’s just in our heads, even if it hasn’t formally reached the schools yet, the fact that it’s there, the potential is aligned and all of a sudden that your friend who knows Shakespeare can really help you navigate the complexities of a relationship with a tool that speaks human language is awesome to me.
Something that has jumped out at me in recent months is the fact that I have an education degree has been so helpful. I understand on coding and technological specifications. And that’s valuable too. You can’t really have one without the other if you’re truly working in the space.
But I can deploy my education degree to know when is the right time to talk about an advanced concept or not. I can remember things like, you’re better off having three main points. I could go all day long about GEO or anything like that with a client, but because I’m so deep in the material, it’s hard for me to remember that most people aren’t. The education degree has helped that. It’s also helped in writing agent level scripts or prompts that actually can work.
Listen to the full interview, with bonus content, here
0 Comments