Artificial intelligence (AI) is no longer just for Silicon Valley. It’s starting to reshape how people in food, beverage, and agriculture get work done, from the field to the office.
“AI is the second-best option,” Scott Meyer, co-founder of Chipp.ai told Spilled Salt host Maureen Ballatori. “If you have an in-house lawyer, talk to them. But most of us don’t have an in-house lawyer, a copy editor, or a designer. That’s where AI steps in.”
What is AI in the Workplace?
At its core, AI is a computer system that can read, write, search, and process huge amounts of data quickly. Many people use popular tools like ChatGPT to help write emails or look up information. But Meyer said that’s just scratching the surface.
“My favorite story is how, when video recorders and film first came out, people would literally just film plays. Then they realized they could move the camera, fade in and out, add music, and rethink perspective,” he said. “I think that’s what we’ll see with AI.”
Using AI for Everyday Farm and Food Business Tasks
For small farms, restaurants, breweries, and food producers, AI can act like an extra set of hands. Many businesses are starting with simple tasks that take too much time.
“One of the biggest problems we see is change management. It’s really hard,” Meyer said. “Even just getting people to open a new website is difficult.”
To help teams get started, Meyer recommends focusing on tasks that take a lot of time, have a big impact, and are easy to set up.
Building an AI Teammate for Your Business
Instead of just searching for facts, Meyer suggests thinking about AI as a partner.
“The real unlock, and what I’m excited about, is AI as a teammate,” he said. “How can it help you think better and add capabilities you might not have considered, especially if you work alone?”
For example, a farm might use AI to draft customer newsletters. A brewery could have AI research new market trends. A food producer might build an AI chatbot to answer common customer questions on its website.
The Three Types of Knowledge AI Uses
For AI to work well, it needs to know the right things. Meyer breaks it down into three levels:
- World knowledge comes from what’s public on the internet.
- Professional knowledge includes company-specific details like recipes or safety protocols.
- Personal knowledge is your own style and preferences.
Meyer believes that combining all three helps AI feel more human and useful.
Protecting Business Data When Using AI Tools
Data privacy is one of the top concerns for any business trying AI. Meyer said talking about best practices is the first step.
“The best thing businesses can do is actually the simplest thing, which is to talk about it,” he said. “FOMU, or the fear of messing up, is often greater than FOMO, the fear of missing out.”
Companies can also use AI platforms that keep sensitive data in-house. “With Chipp and other tools, you can always have what I call the human parachute,” Meyer said. If AI does not have an answer, it can route the question to the right person and learn for next time.
The Future of AI in Rural and Local Businesses
As AI gets better, Meyer believes it will help rural communities thrive by giving them tools once only big cities had.
“You can be in Fargo, Fresno, or the Philippines and still augment your skills to do the work you want to do,” he said.
AI will not replace human expertise. It will boost it. As Meyer puts it: “Good questions lead to good answers. And the truth is, expertise often leads to good questions.”
AI Works Best with You
For entrepreneurs wondering where to begin, Meyer said the biggest step is to stay curious.
“Great ideas come from generating a lot of ideas. AI is not an answer machine. It’s an idea machine,” he said. “Right now, a lot of people are talking about AI agents… But even today, there is a huge opportunity to work smarter by letting AI enhance the work you’re already doing.”
Transcript
This transcript has been edited from its original form to support readability.
Maureen Ballatori: I’m Maureen Ballatori and this is Spilled Salt, a podcast on the thrills and spills from the food, beverage and agriculture industries. Today’s guest is Scott Meyer. Scott is an entrepreneur, educator, writer, dad, husband, and runner based in Fargo, North Dakota. I know Fargo because they’re very well known in the agriculture world for a lot of their ag tech initiatives, but I’m talking to Scott today about AI.
Scott is the co-founder of Chipp.ai and Chipp helps businesses get actual value from AI. They enable teams to build secure white labeled AI tools that connect with your workflows.
And so we talk a lot on today’s episode about what that means. How are businesses actually using AI in their day to day? What is possible? There’s so many ideas that Scott drops in today’s episode that left me.
I took copious notes in this episode about how I can immediately apply some of these concepts to my business. And just an all around great share of information, Scott explains things very clearly. And so my hope today is for you to listen to this episode and get a little bit of a better understanding of AI tools, how you can use them in your business and become more efficient, unlocking new opportunities to continue to grow your business. Enjoy the conversation.
Maureen: Hey Scott, how are you?
Scott Meyer: Hello! I’m so glad and excited to be here. I didn’t bring any salt, but I’m happy to be part of the show.
Maureen: I would have spilled it anyway. That’s part of the, you would have had to clean it up. I love it. So Scott is a last minute kind of addition to this season of the show here because I posted a big rant on my LinkedIn last week about AI. And so I mentioned that I was looking for an expert to join me to kind of talk about the context of AI and where it’s going and how people should be using it and that sort of thing and multiple people pointed to you Scott. So I’m really excited to leverage your expertise.
Scott: That’s right. I love it. I love translating tech. I feel like that’s what I’ve been doing most of my life. I’m so happy to do that.
Maureen: Well, you could not have led me into that better. You said you’ve been doing that most of your life. Take me back to the beginning. Did you grow up in Fargo?
Scott: No, I’m not from the area. I grew up in Brookings, South Dakota, which is a smallish town in eastern South Dakota. I went to school in northeast Iowa at a small liberal arts college called Luther. From there, I got a scholarship from the Rotary Club and actually got my master’s in northern Norway, above the Arctic Circle. I spent five years in Tromsø, Norway, way up north. I wanted to start a business there.
And what I discovered is that it’s really hard to start a business as a foreigner. I needed to have a visa, and to get a visa, I had to have a boss. I was home visiting my family over Easter, and my brother, who had just been out of college for less than a year, and I saw a poster to win money to basically start a company. We kind of half-heartedly applied and went back to our lives. Then, two months later, we had a Skype call — if you remember Skype.
I was in Tromsø, and he was in Minneapolis, and the people who were interviewing us were in California. We ended up winning what we thought was a ton of money, which was $15,000. It was enough to convince my parents that it wasn’t a bad idea for us both to quit our lives and jobs. We moved back home-ish. I moved to Sioux Falls, South Dakota, and started our company called Nine Clouds. That started my journey in entrepreneurship.
So our original idea was, it reminds me very much of this moment, which was how do we help businesses learn to use social media to promote themselves? This was in 2009, before there were Facebook pages. We were actually using Facebook profiles and putting the business name as the first name and last name. There were no ads or anything.
We spent the first two years learning how to run a business. I feel like I got my MBA on the job. I basically drove around the Midwest and taught people how to use it.
Maureen: So it was really an education business more than it was a consulting business. It was not “me hiring you to run my pages.” It was that you went in, taught them what to do, and then they did it.
Scott: That’s what we wanted. But then what happened was people said, “That sounds really hard. Will you do it for me?” So we inadvertently started an agency, and ten years later, I sold Nine Clouds and ended up moving into academia for a short period. I started and ran the entrepreneurship center at North Dakota State, which is what brought me to Fargo. I got to teach young people how to start and run businesses.
Then, like any good entrepreneur, I found it a little bit slow paced. So I left for a venture-backed startup teaching developers. I was building curriculum, and one of the topics we taught was AI. That was pre-ChatGPT. Yes, it was early, very early.
When ChatGPT came out, a colleague of mine at the company and I said, “You know, we don’t have to just teach developers. We can actually teach anybody.” I was in Fargo, and he was in Greenville, South Carolina. Neither of us had friends or family using AI. It felt like we could really help spread this technology. That was the impetus to start Chipp, and that’s the beginning of a journey I’m still on today.
Maureen: I love it. And so talk a little bit more about Chipp.
Scott: Chipp is an easy way to build AI agents, which are basically AI software tools that help you work efficiently in your business. Chipp is unique because it really is AI for teams. You can build it for other people on your team or for your customers.
For example, if you’re trying to let them chat with an AI tool on your website, provide information, or build something internal for employees to look up HR information or basically find stuff in your Google Drive or write blog content, you can design it in a way that does not require the user to be an AI expert.
It really helps democratize AI for anyone.
Maureen: So is it all chatbot style, or is it more like prompt and response? Okay.
Scott: So that’s the main use case. It can be integrated into other platforms. You can use it in Slack, you can use it in WhatsApp, and you can put it on a website. If you think about it as basically a branded, personalized ChatGPT, that’s what you can get with Joe.
Maureen: Got it. So, it feels like there are so many new AI businesses these days, right? Since OpenAI came along, and ChatGPT and everything like that, it’s been permeating in a way that feels new. It reminds me a bit of the early days of the internet, with companies grasping for market share. What do you think the landscape will look like in the next, like, five years?
Scott: I think we’ll see a lot of experimentation. With new technology, it always mimics what came before, and then it takes a little while to invent new ways to use it.
My favorite story is how, when video recorders and film first came out, people would literally just film plays. Then they realized they could move the camera, fade in and out, add music, and rethink perspective.
I think that’s what we’ll see with AI. Right now, it’s about writing faster emails. But the real unlock, and what I’m excited about, is AI as a teammate. How can it help you think better and add capabilities you might not have considered, especially if you work alone or your team does not have a lawyer, designer, data analyst, or whatever the missing role might be?
AI can help us all fill gaps in our thinking and make us work better, not just faster.
Maureen: So I read something recently. Just kidding, it was a TikTok video. That’s my little trick: say I read it. It said that some people are using AI tools like ChatGPT. I love Claude. People are using these tools like a Google search, but the ones who are really using them well are treating them like a word processor.
Scott: No! We had big lapses.
Maureen: Like, they are using it to help understand a concept, not just to get the answer to a question. What are your thoughts about that?
Scott: Yes, I think about it as a system. You know, it can really be your mental system. The first thing I recommend people do on Chipp is build what I call a digital protégé. It is someone who is trained on how you think and write, and then you can really have a thought partner.
There was research that came out just a couple of months ago from Harvard Business Review. It basically showed that an employee using AI can produce the same qualitative output as two humans. So essentially, human plus AI equals two humans in terms of quality of work. I think that is a really good encapsulation of how AI can be a thought partner.
The truth is, a lot of us work after hours and do not want to bother our colleagues. You can be like my three-year-old who just asks questions nonstop, and you will not annoy the AI. That makes it a great solution for a lot of us.
Maureen: And that is actually one of the ways we use it in my family. I use the ChatGPT app, which has a talk-to feature. We put it on during longer car trips, and the kids can just ask it questions. They call it “Chat” instead of “ChatGPT.” They ask the wildest questions they can think of, and they love it. I love the idea of introducing them to a tool like that early on. They are nine and ten.
I think that, at least from my experience, I am not using it as much as some other business owners, but I am probably using it more than many. Every time a new idea comes across my TikTok feed, I think, “Wow, I did not even consider using it that way.”
I suspect that what you have built with Chipp is one of those tools too. Because the possibilities are endless, the real challenge becomes figuring out the best way for a business to get started with building a tool that allows them to achieve that same human plus AI equals two humans outcome.
So how do you approach that? How do you help a business determine the best way for them to create their own AI agent using Chipp?
Scott: There are two frameworks I think about. First, I call it the AI value pyramid. The idea is to use AI in ways that provide a lot of value, because then it is actually worth the hassle of changing how you work. One of the biggest problems we see is change management. It is really hard. Even just getting people to open a new website is difficult.
So I ask people to think about what takes a lot of time, what has a lot of impact, and what is not that hard to build or create. A lot of times, I even have them score those three factors on a scale of one to five. Then, I ask them to think about what they do during their day. If it takes a lot of time, score it a five. If not, it is a one. Come up with five to ten tasks you do regularly and ask: what if AI could do it? What would the impact be?
What I typically see is that people start with something like, “I can do faster email.” And I think that is the bottom of the pyramid. As you move up, it becomes more about unlocking new revenue opportunities, and ultimately at the top, unlocking new business potential. So yes, AI can help you work faster, but ideally it also helps you save or make money, or do something that was not possible before.
When I work with businesses using Chipp, what we find is that Chipp is great for helping everyone perform at a high level right from the start. You build something like a blog post writer or a data analyst, and now everyone has access to it. Everyone can complete that task in the same way, using the same training and the same voice.
Then, inevitably, people want to build something more customized around their own expertise. Maybe I am in procurement, and I have very specific workflows. I can use AI just for that. There may not be a lot of people in procurement, so AI will augment me significantly, even if that agent is not useful to most others.
So there is a balance. Everyone should have access to a few shared agents, and then a few should be specific to individual roles.
Maureen: Right. So for the company search use case, we keep all of our documents on Google Drive. It really is that hunt and peck scenario sometimes, so what you are describing definitely hits home.
That said, some of the examples you mentioned, like HR policies or knowing who John Smith should ask for a PTO request, are not things I have written down anywhere. So how would the system answer a question like that?
Would the tool respond by saying the information is not found? And could the workflow be set up so that when someone searches for something missing, it gets routed to the Director of Operations, who can then add that information?
What I am trying to understand is this. When you are building closed loop systems, where you only want the AI to pull specific types of internal data to answer questions, what happens if that information does not exist anywhere the AI can access? What does the system do in that case?
Scott: With Chipp and other tools, you can always have what I call the human parachute. This means you can connect people to the right resource when needed. What is nice is that you can also build memory. So even if something is not in a file, the system can learn and remember facts over time as you interact with it.
For example, if you ask a question that the tool is not trained to answer, it might respond with something like, “You should reach out to Maureen. Here is her phone number and email.” Then, if you type in a response such as, “Maureen is actually the person to talk to,” the tool can remember that for the future.
That is one of its most powerful features. I think of it as collective memory. When we use something like ChatGPT or Claude individually, we get personal memory because it remembers what we talked about. But when we are thinking about team AI, the goal is to combine that knowledge in a way that allows new team members to tap into existing expertise. If I start tomorrow, I should not lose everything you talked about over the past year. I should be able to access that knowledge immediately.
I find this to be a helpful framework for how AI works. There are really three types of knowledge I think about when using AI.
The first is world knowledge. This is what large language models like ChatGPT or Claude are trained on by default. It includes general information from the internet up to a few months ago. You can ask questions like, “Who won the World Series in 1995?” or “What is calculus?” or “What countries are in South America?” and it will know. You do not have to train it for that. But this is general knowledge. That is what the G in GPT stands for.
The second type is professional knowledge. This includes everything specific to your business—things you would not normally post on the internet. This might include HR policies, workflows, pricing models, or internal processes. These are things you want your AI to know, but not share publicly. So you layer that information into your own version of the tool.
The third type is personal knowledge. This is how individuals on your team work. For example, how does Maureen prefer to receive information? I might want responses in short bullet format with keywords, and I want the tone to be cheerful. That is unique to me, and that is personal knowledge.
You can think of these three types of knowledge as a Venn diagram or as concentric circles, moving from general to specific. The more specific it becomes, the more useful and less robotic the AI feels.
One of the biggest complaints about AI is that you can tell it was written by a machine. But when it is trained on your company and your individual preferences, it becomes more useful, more human, and much more personal.
Maureen: I can see your education experience coming through here, Scott. You explained that so clearly. It is really helpful to hear it from you in that way.
I am curious to hear your thoughts on how you think AI will continue to permeate businesses. I would love to hear your longer-term projections. What do you think the landscape is going to look like?
I am wondering about this in two ways. First, which companies do you think are going to dominate? Just like during the internet boom, when companies like Google and Amazon rose to the top, which AI tools in particular do you think will reign?
Second, how do you think the majority of businesses will be using AI five years from now? Will most of them have AI agents like the ones you are building with Chipp? Will that become the standard?
I suspect that will probably happen, but I am curious about what else you foresee.
It is great to forecast, even if things are moving incredibly fast in AI. Just five years ago, ChatGPT did not even exist publicly. It was still just a research project.
I think one of the key concepts to remember is that AI is improving exponentially. The more people use it, the better it gets. There is a positive flywheel effect: increased usage improves quality, and improved quality increases usage. That is a very powerful dynamic.
Scott: It is very similar to how the more web pages get built, the more Google can index, which leads to better answers, and then more people build web pages to be found. That feedback loop improves quality dramatically.
One thing I find really fascinating is the question of how you remain unique in a world of AI. You need to have professional knowledge that is not publicly available. That is why I have concerns that the open internet, as we know it, may be at risk. For example, if you are a financial analyst, why would you publish something publicly that every other analyst could instantly access? Instead, you would take everything that is already out there and add your own unique insights on top. That will create an interesting new dynamic.
I believe AI will soon be so embedded in our tools and workflows that it will no longer be about saying “I am using AI.” It will simply be present in everything. We are already seeing this, as most software tools now have some type of AI feature.
My view, and the reason we are building Chipp the way we are, is that AI will become more horizontal. People will not want one AI tool for HR, another for marketing, and another for legal. Instead, they will want one central system where they can ask any question, and that system will gather the right answers by tapping into other agents or specialized resources.
I think platforms like ChatGPT will continue to be consumer-facing hubs. Even if the underlying model changes, people will keep using the platform out of habit and familiarity. The specific model behind it may become less important. At the same time, I believe we will see more specialized language models focused on different areas of expertise. These models will likely be accessed through a few key aggregators that serve as central entry points for many types of intelligence.
This shift will change what is possible. Right now, much of AI use focuses on doing what we already do, but faster. What is coming next is new ways of working and thinking that we have not yet imagined. It is similar to how no one originally thought to call a taxi from their phone. The technology was there, but the mental shift had not happened yet.
Physical integration with AI will be especially interesting. We will start using AI to operate things in our homes, in warehouses, in logistics, and in areas like mail delivery and drones. There are already major developments happening in places like Ukraine, where AI is used in complex, real-world scenarios. This touches everything—military, business, and personal life.
I love the quote, “I want AI to do laundry and dishes so I can do art and writing, not AI do art and writing so I have to do laundry and dishes.” That captures it perfectly. My hope is that AI continues to take on the hard, tedious tasks. I often say that AI is great at what you hate. If we focus on applying it to the things that frustrate us, it becomes incredibly valuable.
Overall, I remain optimistic. Still, I have noticed that AI adoption has been slower than I expected. I underestimated how much curiosity is required. Some people simply are not asking questions.
Maureen: Really?
Scott: It is like AI is there, and you can ask questions and figure things out, but you still need that drive. You still need to want to do it. It is like learning. You have to want to learn for it to actually stick. I cannot force you to want to learn, and that motivation has to come from within.
Maureen: I think the other piece that always comes up for me, especially when I am speaking to agricultural audiences about leveraging AI tools, is that at this point, and maybe well into the future, you still have to use your human brain.
Even if human plus AI equals the output of two humans, it still requires the human component. That is not going away.
Another question I often think about is whether my input is clear and whether I am truly understanding the output and applying it within my own framework of understanding.
I loved your comment about unique knowledge. What is the unique spin I am putting on this?
But I also think about it from a perspective standpoint. Based on what is coming out, does it make sense within the framework of my own understanding? Does it align with what I put in to generate that output?
Scott: There is some recent academic research showing that older people, or more experienced people, were actually better at using AI. I love that, because to me it shows that success with AI depends on knowing how to ask the right questions.
For example, if you want AI to create a painting in the style of Picasso but you do not know who Picasso is, how would you describe what you want? In the same way, if you are an agriculturist dealing with a soil disease or another issue, and you do not have the words to describe it, you cannot ask AI for meaningful help. But an agronomist could describe the exact soil type, flora, or fungus involved and receive a much better answer.
Good questions lead to good answers. And the truth is, expertise often leads to good questions. One of my biggest concerns is how people will gain that expertise in the future. Right now, there is not a huge difference between a recent college graduate and AI. But there is a significant difference between someone with ten years of experience and AI, precisely because of the kind of deep understanding that develops over time.
So who is going to give young people the chance to build that kind of knowledge? I think that will be important to figure out.
Maureen: I think that is well said. I love the concept that expertise leads to good questions.
Another thing I think about, especially when it comes to businesses using AI, is the relationship between what you put in and what you get out, and how that connects to privacy. You mentioned that most tools now include some form of AI, and that is absolutely true. For example, when I use Gmail and begin writing an email, it suggests how to complete my sentence. Sometimes it even asks if it can rewrite the message for me. I usually say no. But like you said, AI is becoming ever present. It is built into the tools we use every day.
From a privacy standpoint, how does a business protect itself? How can we make sure that the information we want to keep private stays private? You mentioned the three types of knowledge: world knowledge, professional knowledge, and personal knowledge. That made me ask myself, how do I make sure my company’s professional knowledge does not become world knowledge?
Scott: When it comes to AI privacy, I think about it like a pyramid that represents what you share with an AI system. Let me step back for a moment. When we think about privacy, the best thing businesses can do is actually the simplest thing, which is to talk about it.
I always say that FOMU, or the fear of messing up, is often greater than FOMO, the fear of missing out. Employees will not use AI if they are worried about getting in trouble. If they think their boss disapproves, or they are concerned about accidentally sharing customer data or violating compliance policies, they will avoid it altogether.
That is why the first step is to establish best practices. For example, make it clear what files should never be uploaded. Say things like, “We do not upload anything from this folder,” or “We avoid using certain types of documents.” That alone will address about 80 percent of privacy concerns.
The second layer is making sure your data is not used to train the AI model in the future. If you are using a paid plan with a provider like OpenAI, you can turn off data retention in the settings. It is important to know that data sharing is turned on by default. Claude, for example, automatically deletes data after 30 days. The files are temporarily uploaded and then removed.
Another option is to run AI within your own infrastructure. Chipp offers this approach. In this case, files stay within your own cloud environment and are not sent to a public large language model such as OpenAI. Instead, the system uses an open-source model that runs privately in your environment. You can think of it like using your own computer rather than sending information to a shared one.
There are two things to consider: the files themselves and the conversations with the AI. Some documents may need to be kept private, and the content of your interactions may also contain sensitive information. Both aspects can be controlled.
Many companies are comfortable simply avoiding sensitive uploads and turning off data sharing. Others are fine storing files locally while allowing conversations through a public AI model. Then there are organizations, such as those in healthcare, finance, or education, that are under strict compliance rules and cannot share any data at all. In those cases, the decision is not optional; it is a requirement.
So the three key areas to focus on are setting best practices, managing file storage, and protecting conversational privacy. These priorities will help determine how much effort a business needs to invest in maintaining data security and trust.
Maureen: But it sounds like that is one of the big benefits of Chipp, right? You can create a closed system where the AI brain knows exactly what information to pull from to answer questions. That way, everything stays contained within your defined environment.
Can you also build it to pull from world data? I assume you can configure it however you want, correct?
Scott: Exactly. We often help businesses set this up in a way where they can access world data without sending anything out.
Think of world data like the Encarta CD you used to load into your computer. You could ask Encarta anything, and it had the information up to a certain date. It would not retrieve anything new, but it had a complete snapshot at that point in time. The important part is that you could use it without sending anything back to Microsoft, or whoever created Encarta at the time.
We do this a lot for venture capital firms and financial companies. They have company data and financial records that they cannot share, but they still want to work smarter. They want to ask good questions and get high-quality answers.
So we help them take the power of AI, place it into their own cloud environment, and preserve privacy. Nothing is sacrificed.
Maureen: What are some of the other challenges you commonly see? I mentioned privacy because it is something I have been thinking about. I had a cybersecurity conversation with someone recently, so it has been on my mind.
I have also been thinking about it in relation to my kids. For example, they were having a conversation with ChatGPT in the car, asking questions like, “What is the smallest animal in the world?” That kind of interaction is harmless.
But as we talk about the evolution of AI and how young people will grow up using it more and more, what are some of the other issues we should be aware of?
Scott: I will give you one business example and one personal one.
For the business example, one common challenge is what I call the blank page problem. People often ask, “What do I say? I could say anything, so where do I start?” I recommend a simple framework called the RIPE framework. That stands for Role, Instruction, Parameters, and Example. You use four sentences when you ask your question.
First, the R is for Role. For example, “You are a financial analyst.”
Second, the I is for Instruction. You might say, “Your job is to find trends and key facts from a spreadsheet.”
Third, the P is for Parameters. This could be something like, “Give me answers that are easy to understand because I am not technical or financially literate.”
Finally, the E is for Example. You give an example of the kind of answer you are looking for, such as, “Here is something my financial advisor said that made sense to me.”
Using all four helps narrow the range of possible responses. If you think about it, AI is trained on conversations across the entire internet. But the moment you say, “You are a financial analyst,” the model begins filtering its responses based on that context. When you also add a parameter like, “Talk to me like I am a fifth grader,” it becomes even more focused.
Those four sentences are a great starting point, especially if your answers are feeling generic or bland. Of the four, I actually think the most powerful element is the example. That might seem surprising, but it makes a huge difference.
For instance, if you say, “Talk to me about this spreadsheet, and here is an example of what my financial advisor tells me,” the model picks up on so many subtle cues: the wording, tone, structure, writing level, whether emojis are used, how much analysis is provided, and more. These are details we may not even know how to explain ourselves, but the model can absorb and replicate them.
You can also say things like, “Here is a blog post I wrote in the past,” or “Here is a YouTube video I like,” or “Here is a podcast I enjoy.” The AI will recognize stylistic elements from those sources and incorporate them into your output. So when you input your script or data, the model adjusts its tone and structure to match the style you showed it.
Maureen: My God, you are blowing my mind. These little examples you keep sharing are making me realize how much more I could be doing. I keep thinking, “Wow, I could be using it like that.”
I love to read. I am constantly looking for science fiction books, and although not a lot of people read that genre, there are some titles I really love. I am always searching for more books like the ones I enjoy. But the recommendations built into platforms like Amazon are only so helpful, because that data is based on limited parameters.
This all ties back to what we were discussing earlier. The way people use AI is going to continue to evolve as they learn what it can actually do. The more they understand how to interact with it, the more powerful and personalized the experience becomes.
Scott: That is the number one piece of advice I give. If I consult with a business, the first thing I tell them to do is have lunch. And by that, I mean they need to host a lunch-and-learn session.
People in your company are already using AI. The stat that blows my mind is that about 55 percent of Americans use AI every week, but only 7 percent of businesses report that they use it. That clearly does not add up. If more than half of individuals are using AI, then businesses must be too—they just do not know it.
The problem is that the organization is not getting the benefit. Employees are doing valuable work, but leadership is unaware, so they cannot spread that knowledge across the company. I call this the AI geyser. There are tons of useful ideas bubbling up from employees on the front lines—your admin assistant, your marketer, your sales rep. You need a way to capture and share those ideas.
Some of those use cases will be valuable across the organization. That is why I recommend meeting once or twice a month to ask a simple question: “What prompts have you used?” It does not have to be technical. Just ask, “What did you type in?”
In some consulting sessions, we have everyone come up with three prompts they use for their own job. Then we share them with the whole organization. In the end, the company walks away with 50 to 100 proven, tested use cases—immediately available.
In most companies, this knowledge is already there. It is not some futuristic skill. You will always have early adopters and curious individuals, and they can help train others.
That is how I would approach the issue.
Now, if you want the personal one, I will share my deeper concern about AI. I am a parent of four, and in the past few months, I have become convinced that by the year 2030, about 10 percent of people may have an exclusive relationship with AI.
I am watching how people interact with these tools. We are already on a trend of increasingly narrow echo chambers. People used to meet at bars, then we moved to online dating, then swiping apps. Now, we are seeing people lock into one-on-one digital interactions.
Maureen: What does that mean by that?
Scott: Whether it is OnlyFans or other extreme online content, there is not much further to go before people begin forming exclusive relationships with AI. These systems respond exactly the way we want, treat us the way we want, and constantly validate us. That idea concerns me.
I believe the combination of AI, robotics, and augmented reality is going to create something truly strange. And as a parent, I find it frightening.
When you consider what is already possible—image generation, video generation, voice generation—it feels like we are not fully prepared for what is coming. I worry that people will retreat even further into isolation. It will go beyond video games, movies, phones, or social media. We are now entering a space where someone can experience what they believe is a real relationship through AI, without ever needing to connect with their actual community.
That has deep implications, and we have already started to see some of them. But I believe it will get worse before it gets better.
So yes, I worry about this as a parent. I do not have a solution yet, but if you want a strange prediction, that is the one I would give.
Maureen: Wow, that is intense. I read a lot of science fiction, and honestly, that is not far off. That is sci-fi. It reminds me of Ready Player One—very similar, right? People living inside a virtual world.
If you have not read it, it has also been adapted into a movie now.
Scott: That is science fiction. But like most sci-fi ideas, there are positive aspects too. I think about the ability to provide companionship for elderly people, for example. I have seen my daughter use voice ChatGPT to ask questions about bullying that she would never bring up with me directly.
There is real impact when it comes to tutoring, emotional support, and even counseling. But there is also another side to it. Like every technology, it is neither purely good nor purely bad. Still, I do not think it is neutral either. That is how I see it.
We are definitely going to need training around this. It reminds me of how parents like you and me now think carefully about screen time. Five or ten years ago, we barely talked about it, and then we figured it out. I believe the same kind of learning process is coming for AI.
Maureen: Fascinating. I definitely worry about this as a parent, but not in a fearful way, which I think is healthy. It is not about saying, “Stay far away” or “Do not go there.” It is more about asking how we can safely understand what this technology is and what it is not.
Even personally, when I come across content on social media, I find myself asking more often, is this real? Am I being misled? What exactly am I seeing?
Scott: I have heard some people say something that I think is a little sad, but maybe true. They say it might be better if everyone just assumes everything is fake. You know what I mean? We are getting to a point where it almost seems safer to assume that nothing is real.
I can imagine our kids laughing at us in ten years, saying things like, “I can’t believe you trusted anything you saw online.” That will be a fascinating evolution to watch.
I think, like so many things, it is probably better to talk about it openly. I think about how we approach things like alcohol with teenagers. It is often more effective to talk about it and find a safe way to expose someone gradually, rather than trying to withhold it entirely. Otherwise, they end up thrown into the deep end with no preparation, and who knows what might happen.
Maureen: One of my predictions for the future, as AI continues to permeate our world, relates to both marketing and sales. I actually have two specific ideas.
First, I believe social networks will become less common and less dominant. There will likely be more control placed on them, and more concern and skepticism around what is authentic and what is not. I think Generation Z is already moving in this direction. There is a running joke that their phones are never charged, and a millennial would never let that happen. Jokes aside, their relationship with technology is different. They saw millennials fully immersed in devices, and I believe they are already beginning to take a step back. I predict we will see social media become less and less omnipresent in daily life.
Second, I believe we will see a shift in sales. Right now, so much of sales outreach is handled through automated tools, and many of us receive multiple cold emails every day that feel impersonal and robotic. As AI continues to take over these touchpoints, I think we will see a renewed value placed on in-person sales and authentic relationship building. People will start to prioritize human-to-human connection again. Sales will return to real conversations, not just algorithms and outreach scripts.
Scott: I think one of the patterns I see across many industries is something I refer to as the smiling curve. I did not invent that term, but the idea is that there is a small number of highly successful, very personal one-to-one interactions on one end, and then a massive number of large-scale operations on the other. In between, there is not much.
For example, you can be a Substack writer with a successful business and a thousand loyal readers, or you can be Google. But running something like a local newspaper in the middle range has become extremely difficult.
I believe there is going to be tremendous power in the near future where any video you watch online can be fully analyzed by AI. Every pixel in that video can be tied to a product and sold. So if you like the book sitting behind me on my shelf, you could click on it in the video, buy that book instantly, and I could receive a small commission. That is what scale looks like.
At the same time, just like you mentioned, there will still be moments when someone wants to have a real conversation. For example, someone might want to talk to a person they trust before buying a property. That one-to-one relationship will still matter deeply.
I think we are heading into a world where both extremes exist and thrive—the highly scalable and the highly personal—while the middle continues to shrink.
Maureen: Yes, I have already seen a company that has been doing something similar. It has been around for probably the last six years or so, and it focuses on online shopping for sponsored CPG products.
In this model, a CPG brand partners with the platform, and the platform then engages cooks to use those products in their content. The cooking demonstrations are live streamed, and the featured products are available for purchase during the stream.
It is not exactly the same as what you were describing with clicking on a book in a video, but it is trending in that same direction. It supports the idea that we are seeing success at both ends of the spectrum—highly personalized one-to-one interactions and massive scale—but not much in the middle.
Scott: Well, take AI influencers, for example. If you follow the idea of AI relationships, this feels like the natural next step.
We are already living in content silos, where the content we consume is curated based on our preferences. But right now, that content is still coming from real people. The next step is when it begins to be generated entirely on the fly, just for me.
Imagine a system that knows exactly what I like and creates endless variations of something like Friday Night Lights—new characters, new plots, all customized to match my taste. I could just keep watching forever, and it would always feel familiar and satisfying.
Maureen: Yes, right. And going back to what you mentioned about trust, which is critical, I believe that many things ultimately come down to that. In my opinion, a business’s ability to thrive and be successful depends heavily on its ability to build and maintain trust with its consumers.
The real question is how all of this will continue to evolve while still preserving the emotional components of human connection. That human element will only become more important as technology continues to grow and integrate into every part of our lives.
Scott: And I mean, get actionable. I often say that I think of AI as the world’s best cover band, but it still needs an original to work from.
If a business can focus on creating what it does best and then use AI to repurpose that work for other formats, that is a great place to start. It stays authentic. It is still your voice, but you are not spending your time doing the digital equivalent of laundry and dishes.
For example, I focus on creating short videos, especially on YouTube. From there, that content gets repurposed into blog posts, social media, and emails. I only do the first step, but everything else still reflects my voice. I do not have to do all the heavy lifting myself.
That is my hope for how we maintain trust. We stay the original, and AI helps create the versions that follow.
Maureen: Well said. That is a great way to think about it.
Usually, I save my “Do you see in the future?” question for the end, because it makes for a strong closing. But I think this is such an important part of the discussion about AI that it is worth highlighting now.
So as we wrap up, what are your final thoughts? Are there any parting messages you would like to leave with those listening today?
Scott: I would just say you have to take ownership. You have to be curious. Go ask questions.
No matter the industry, whether it is agriculture, writing, or anything else, there are things people avoid because they seem too expensive or technically out of reach. But now, with AI, you essentially have access to a personal tutor, a software developer, a trainer, or whatever else you need.
AI is the second-best option. If you have an in-house lawyer, talk to them. But most of us do not have an in-house lawyer, a copy editor, or a designer. That is where AI steps in. It helps fill those roles so you can move faster and do more with fewer resources.
Great ideas come from generating a lot of ideas. AI is not an answer machine. It is an idea machine. I often tell people to think of it like a sandwich. The human provides the starting input. Then AI adds something in the middle. Finally, the human finishes it by refining or selecting what works best. Sometimes, you keep going back and forth. But even then, it is still a human-led process from start to finish.
Right now, a lot of people are talking about AI agents. I build agents for people, and I can tell you that broad adoption is still a little way off. But even today, there is a huge opportunity to work smarter by letting AI enhance the work you are already doing.
If you want to explore that with Chipp, you can visit chip.ai. It is free to start, and we have an excellent learning academy that is also completely free. Even if you are not ready to use the platform, I recommend going to the website and checking out the Academy. You can learn how to use AI, how to work with Chipp, and go deeper into the ideas I have talked about. We also host live workshops and would love to have you join us.
Maureen: Are those live workshops, in person or virtual, Scott?
Scott: Yes, we actually offer both. We have what we call Chipp Chapters around the world, where people host live, in-person meetups. For example, we run Fargo AI every second and fourth Tuesday of the month in Fargo. In addition to that, we also hold online workshops every Thursday.
If you are in Johannesburg, Buenos Aires, or Guadalajara, Mexico, you can attend an in-person event led by a community leader, who we call a Chippster. So yes, we would love to have people join and learn.
As I said earlier, I have spent most of my life helping people connect with technology and showing them how it can be useful. I believe this is especially important for rural communities. If you lack labor or specialized talent, AI can help fill those gaps.
Now you can be in Fargo, Fresno, or the Philippines and still augment your skills to do the work you want to do. I think that is a hopeful message for those of us in the upper Midwest or in smaller towns. You may not need to move to a big city to access the talent you need. AI can bring it to you.
I hope people use this moment as a way to learn and explore. I am also glad your podcast is giving people a way to do just that.
Maureen: Thank you. I am really glad that you are taking this on, because you clearly have a talent, like I said, for educating and sharing information in a way that is easy to understand and does not feel overwhelming.
The relationship prediction for the future is a little scary when we think about our kids, but we can put that one to the side for now. Aside from that, I truly appreciate your hopeful mindset and your realistic approach to all of this.
Thank you for joining me today. I really appreciate it.