AI tools generate quick spreadsheets and answer basic operational queries, but relying on software for AI Business Advice leaves major strategic gaps. In this conversation, Gwen Bortner and Tonya Kubo explore why hiring experienced human experts remains vital in an era of automated information. While algorithms process vast amounts of data, they lack the specific context and practical intuition required to guide high-stakes business decisions. Tonya and Gwen unpack how software often surfaces textbook solutions rather than tailored action plans for unique operational challenges. They also discuss why relying heavily on automated tools risks weakening a founder’s strategic executive judgment over time. Discover how to leverage technology while protecting the human connection and critical thinking that fuel real business growth.
—
Watch the episode here
Listen to the podcast here
Why Pay For An Expert When AI Can Give You An Answer?
I cannot say that we’re going to give you a concrete, perfected-every-scenario answer, but that is the question, who?
They could run it through AI and see what it says.
Information Vs Strategic Guidance
Their AI could talk to our AI, and our AI could respond, and it could just be an entire AI field day. Gwen, something that you and I have been talking about is that AI has really changed the value of information. Information used to be scarce, used to be hard to find even when you’re doing Google searches, and it was not always easy to type in the right terms to get the information you wanted. Is there, would you say, advice that you used to get paid to provide that you now think can be easily answered by AI?
For sure. Back to I’m a dinosaur. I was doing this work to some degree before the internet was really accessible in the way that it is now. I watched it become accessible over time and become more and more accessible over time and have really seen the search before there were search engines. I’m ancient. Watching search engines start to do their thing and watching the large language models start making it easier in all of those issues. There are a ton of things that, at some point historically, I would be giving advice on.
I’ll take a super simple example is someone wanted a time management system. I had been using and playing with quite a few of them, and I could talk about why you might choose Trello over Asana, over Todoist. Those are just three that popped into my head. There’s obviously a lot more. That would save them quite a bit of time, as now asking me that is just a foolish question because, first off, I do not have the most current information because they’re changing all the time. AI can do a much better detailed comparison. If you give it really good parameters, it will really focus on the pieces that you need to make a better decision.
AI is also dated. It’s not completely current as well. That is an area that people forget. Actually, I was trying to come up with the right executive function app for Brian because he’s just been really struggling. I thought I knew a whole bunch of them, but there’s probably a bunch I do not know about. I’ll take that to AI, and I will ask the question. One of the things that Claude came back with is that, because I have it to tell me, as I’ve programmed it, you need to tell me when you’re making assumptions versus when you know something to be fact.
AI can compare feature lists in seconds, but it can't evaluate the unique operational trade-offs of your business. Share on XIt says, “You should know that my information only goes as far as 2023. If any apps have been invented since 2023, they are not here.” Yes, and sometimes. What other things do you think there are? I know you work with a lot of clients who use AI and love AI. What are some things that you see that they’re using that you think maybe they would have asked you about before?
A lot of them are asking about processes. They’re asking AI about processes that they used to ask me about. Often, quite honestly, it’s giving a better answer, a more complete answer, a more detailed answer than I would have ever been able to give in any cost-effective way, back to depending on how well educated your particular AI is. It’s got enough context to actually do a better job with it.
That’s another place that it’s asking, it’s providing services that maybe I have provided in the past. Now, another one, if we go back to my old life, not my most recent but my old life doing programming and doing a whole lot of that stuff way, way faster. That being said, I’m working with a client who’s working with a programmer programming team, and they very specifically still code in the old way.
They know how to use AI, and they’re leveraging AI in the reporting functions and in a whole lot of ways. They also understand that it’s not flawless. They’ve got their own checks and balances. I’m working with a client who’s right now trying to put in some AI to do some things. In some cases, it does it really well. In some cases it’s not. Having the person in there. There are lots of different places where maybe I provided advice before.
I would say a lot of what I’ve done in the past few years is not just about activities. There’s a bit of wisdom and the challenge with AI, not that it cannot do it because I’ve seen it do it, is that it’s got to have enough context to be able to give really good advice. Unless you’ve taken the time to really develop what that context is, then you’re not necessarily getting the good advice from it.
The Limits Of Open Knowledge
What it comes down to is that AI comes preloaded, as it were, because let’s be platform neutral here, it comes preloaded with everything that exists in the open access space, all the information that’s out there. That’s good information. That’s bad information. It is surfacing answers to you based on what you’re asking and the frequency of responses.
Without deep context, AI just regurgitates open internet data. Real advice requires knowing what isn't being said. Share on XAI is not great when you are somebody who has followed a textbook and the textbook solutions do not work for you. That’s hard. Also, you have to have enough knowledge. It’s almost like you’ve got to be a really good consultant in and of yourself to upload the most valuable context of your business to allow AI to be your consultant.
An interesting example that I’ve seen several examples of is the Delphi product, which is a clone of you. People that I know personally who have used it are Mike Michalowicz, Joe Polish, Lisa Smith, and several other people that I know personally who have used it. The places where it works really well, I’m going to pick on Mike Michalowicz and Joe Polish, because they’re both big names and have produced a tremendous amount of content. Not just some.
They’ve produced the majority of it from their own knowledge, experience, words, whatever. They may have leveraged AI and helped trim it or whatnot, but it’s theirs. They’ve been doing this for a long time. When they set up a clone, they are putting all of that in there. This is when AI starts having a lot of really good information because it can take, I do not know how many books either Mike or Joe had written, but it’s 6, 7, 8, 9, 10. It’s a real number, usually across lots of topics. We’ve got any blog posts and all of their podcast recordings. The list goes on and on. AI can actually do a pretty good job of being able to go back and forth and replicate that person.
Most of us do not have that size of a body of work that allows our wisdom, knowledge, and experience to even come close to what an AI would produce. Back to this is a situation where we’re feeding an AI engine to specifically produce something. That’s where the balance is. I happen to know both Mike and Joe personally. I know that the clone is probably right 95%, maybe even 98% of the time. There’s still going to be times where, because they have a relationship with the person, there’s going to be something in there that they would say, “For you, it should have been such and such.”
Features Versus Practical Judgment
The other thing with AI, and this is something that I have been, as our podcast listeners know, I’ve spent most of the last year really simplifying my tech stack. I’m simplified down now to the place where I can go, “What can I replace these things with?” I’m between a couple of ideas. This platform or that platform. I find that AI can do a great job of giving me a features and benefits comparison.
It cannot go beyond that. It cannot really guide me. I can do deep research and get some guidance on what Reddit forums are saying about each of the tools. When it comes to the really persnickety use cases that I would be putting these tools through, AI can give me answers. It cannot give me guidance. How do you see that in your world?
This goes back to, once again, that the context matters because that’s where you’ve got to have enough context to give, as you said, guidance. Back to, I’ve seen AI actually give good guidance, but not quickly. It’s someone who has invested lots of time helping the AI tool, whatever AI tool or tools that they’re using, get lots of context. This is the place that I think is missed a lot is we say, “AI can just do that.” It’s like AI can, but often AI cannot without you investing a tremendous amount of time. I’ve got a couple of AI experts that I’ve been hanging out with lately.
One of them I was talking to, and I said, like, “I know, technically, I probably should have AI, like, going through my email and filtering my email for me.” I said, “The problem is, I only spend about five minutes, maybe less, a day filtering my email.” I may still need to answer it, but it comes down to a few emails. It’s not like I’ve got 8,000 emails, and I do not end up with 4,000 emails in my inbox. It stays pretty tight all the time.
The Context Trap In Automation
I said, “For me, the math I’ve done in my head is the amount of time it would take to explain all of the context in the way that I think, in the way that I’m processing email, is so large that to get the payback would take years and years for me to actually get the payback.” Part of this is because I’ve been tightly managing my email for probably 25 years, because I’ve actually had email for 25 years. I was managing email when no one would have 1,000 emails because there weren’t that many people with email to email you.
That’s not necessarily a good use case in my situation. That’s the other piece that we have to be really careful about. Is this a good use case for you in your particular situation? The email example is one that everyone is saying is like, “If you’re letting AI manage your email, you’re a fool.” Maybe I am a fool, but the couple of AI experts that I actually trust are like, “No, not worth it for you in your situation with your experience and how it works for you.”
It may be worth it if you run a company and your customer service inbox gets 200 emails a day. It might be worth it for them. I do know, however, that I have bought products from organizations that do use AI for their customer service. I will tell you that 4 out of 5 email responses I get actually do not align with the context of the question I asked. I have to go back and forth three times to either get to a human or get the AI right to understand what I’m asking for.
This is the interesting thing back to the conversations I’ve been having recently with my experts is to get really clear on what needs to be human, where, and actually making a thoughtful decision about what is the piece that AI can do in a way that allows the human to do more human things that actually enhances the experience of the customer, of the business, of your life. All of those things. Multiple answers here. That’s the piece that we’re not used to thinking about.
A feature comparison isn't strategic guidance. Know the difference before handing operational decisions to software. Share on XOne of the gentlemen I’m working with, it’s the question, I told him, “I want to meet with you regularly because what I want to have is the conversation I want to have.” I said part of me knows I’m like, “Get off my lawn, you whippersnapper.” I know that’s not smart. That’s just not a smart way to approach it. At the same point, it means I need to think about things differently. That’s a big leap to think about what actually could and should be AI versus what actually should remain human.
How do you think about it and those judgments? Where does it go? When am I being a crotchety old woman and fighting it, as opposed to when am I actually saying, “No, I really think this still needs to be human?” It’s a really interesting balance that we’re in. The biggest challenge I think that we’re going to have with it is that it’s all changing super fast.
It’s partially why it is a lot of why, because I’m the person who plans out our episodes for the most part. It’s why we haven’t had an episode about AI yet. It’s why I really hesitate to have one.
It’s because it will not hold up.
It’s why I hesitate to even have one now. By the time this comes out, so much of what we say may be already outdated. It may not be. We have no way of knowing because we do not control.
If spending hours training AI takes more effort than doing the task, you're losing money on automation. Share on XWhich is why I’m trying to talk about it as generically and principle-based as possible because the principles will apply.
Curating Focus Over Endless Options
Following that principle-based lead that you have given us is like when I was saying before. An answer is not guidance. A lot of what a human can do that is hard for AI to do is recognize when you’re answering, when you’re asking the wrong question, and also curating your options. I know one of the things that frustrates me when I’m trying to go into AI and just have a very narrow conversation is that it always wants to give me five more new ideas. It’s like, “We could do this, or we could do that.” If I say, “That sounds interesting, but I want to get back to this,” the next response is five new ideas. Whereas when I hire an expert to solve a problem, we stay very focused on solving that problem.
Why Pattern Recognition Misses Intuition
Usually. Let me give ourselves some reality check too, because we could do the same thing. What you’re saying is true is AI is a weird blend of a very tactical, practical advisor and a visionary at the same time. As you said, if you’re not careful, it can take you down a rabbit hole. One of the things that’s also interesting is what people will say, but it’s really good at pattern recognition. AI is good. AI is technically better at pattern recognition than we are. However, it’s only based on what it has seen, either through the broad lens of the universe or if you constrain it to what into your own universe, what it has seen that you have shared with it.
It does not necessarily understand the part of the pattern recognition that may need to go in there, that happened at the soccer game while you were watching your kids, and what happened as you were driving home, and the thought that you had just before you fell asleep, or the idea that came to you while you were in the shower. Technically, AI is better at pattern recognition than we are. It’s the nature of it. That’s how it develops, how it’s enhanced, and continues to grow is via pattern recognition. It’s only the pattern recognition that it has access to.
Some other areas where I see. It’s funny because sometimes five stars go to AI, and then other times it’s like one star. Other areas I think, are judgment. Knowing what information matters. It’s like you type in this whole life story into AI and it says, “This changes everything. What especially calls out to me is this one thing.” You’re like, “That’s actually not the point I was trying to get across.” It’s funny, as I programmed at the advice of somebody else, I had programmed my Claude to challenge my assumptions, and every answer would start with, I’d like to push back on these assumptions and it would list three assumptions that were not assumptions I was making.
AI gives you five new ideas when you need one solution. Great experts curate options and keep you focused. Share on XI’m like, “You just wasted so much of my time.” Part of what experts do for us is curate information. They help us focus our thinking. I know what I value in you, particularly, is your ability to recognize what’s not being said. What is usually part of this conversation that’s not part of this conversation now, based on your experience? What that means is that a lot of times AI can give you a technically correct answer, but it’s still a really bad business decision for you.
It is the other thing. It’s body language. That you may or may not be seen, depending on who your advisor is and how they’re communicating with you. It can be the actual accountability because one of the things is that we still know it’s a computer. Sometimes just having someone that you know is an actual human say, “I’ve noticed that you’ve pushed this off.” Now, potentially, AI could do the exact same thing. It’s like, “You’re just a computer.”
I’m asking you in a conversation like we’re having right now, but I’ve noticed you push this off. What do you think is causing that? You might give the exact same answer, but you also might not. I’m going to watch you. I’m going to watch how you respond. I’m going to see if that makes you jump back and say, “I’m going to see if you start laughing. I’m going to see a whole lot of things that, at least at this point, let’s say that AI is not actually able to really judge in any way.” Now, as we said, it’s moving just truly at the speed of light.
In a month or two after this comes out, it could be like, “No, it’s watching the entire time. It actually knows what all of those things mean. It may be able to do that piece. There’s still a different thing about having committed to another person that is, in fact, an actual person. At some point, we may have robots and all the things that we have seen in sci-fi that could actually become true. I actually do not think it’s that ridiculous where people are building true bonds with artificial beings. We’ll feel bad in the same way that we feel bad when we let another human down.
We’re not there yet. I’m hoping that the podcast lasts long enough that someone’s like, “We are there now.” That would actually make me excited. We’re not there yet. We’re definitely not there yet. That’s not going to happen in the next month or two. We’re way away from that. That’s the thing. There still is a human connection that comes back from our own evolution that requires a different relationship that we have with a human than we are going to have. Sometimes it’s truly about the actual accountability to another human. Not to a thing, not to a process, not to a checkbox, not to a question, but to another human.
There was this MIT study that I’m trying to pull into this conversation, but it does not really fit. Here’s what I will say, because if I say MIT study, and I can put some people will know what I’m talking about, it has to do with writing. It has to do with using AI. Do you actually have the ability to answer critical questions about what you wrote and what you did not, and it found out that you did not. Anybody who has done writing or any creative work would say, “Duh,” because part of your ability to recall and answer the hows and whys is by doing your creative work from the ground up.
If something else is generating it and you’re just reading through, you do not have the same retention. What I find with AI in a business context, along those lines, is that something that you have always said is that you do not say it quite this explicitly, but do not hire a coach or consultant that cultivates or facilitates dependency. If you hire somebody and their answer is that you have to always work with them in order to be successful, they’re not very good. I almost say they’re not good at their job.
That’s not a situation you want to get into because you want to be able to hire an expert that helps you become better in some way so that you can be better independently. Back to the whole purpose of our show has been helping business owners make better business decisions. Those best experts would help you understand why something matters, help you see things that you’re not seeing, help you evaluate trade-offs, and ultimately strengthen your own judgment. Unfortunately, what I am seeing with the use of AI is that people seem to think that they no longer have to worry about their own judgment because they have AI to fall back on for that.
Strengthening Founder Business Judgment
This gets us to a completely different spot that we can probably go on for hours. No one wants us to do that, let’s be honest. Is that we still need to be making judgment. If we’re not careful, we get out of the making judgment piece. We assume AI has better judgment than we have. I’m going to be honest, in some cases it might. In some cases it might have better judgment than we have. Judgment is also like a muscle. If we’re not using it, we’re going to lose it. This, to me, is the balance and the challenge. It goes back to what I was saying toward the beginning is what actually makes sense for AI and what is the human element?
What is that human element? It is about putting those pieces together in a way that is actually useful for you and what you’re doing. When AI first hit the scene, not very long ago, even though AI has been around for a long time, the large language model is actually the piece that’s really fresh, I’ll say. As I said, if someone’s six months out there like, large language model? That’s been around forever. Some of these things will date themselves very quickly. The whole idea of AI in and of itself is not new. It is about thinking where does that human element bring things that AI cannot?
Algorithms spot data patterns, but they can't replicate human instinct or real-world experience. Share on XInstinct is one of them. Gut feeling is one of them. Those things are real. They are not fake. Maybe someday it’s going to be able to do that, but not today. This is the place where you need to keep your human piece of it. It also is the place that you’re going to potentially find your advisors being able to say, “This does not feel right to me.” That’s the place that you want to be looking at. As AI has started rolling in, I haven’t been that worried about what I do because a lot of what I am doing is about taking all of that experience, all of that knowledge, and mushing it together. Some of it is instinct and being able to ask different questions that AI probably wouldn’t ask, because I focus on what the context is.
As AI keeps getting better, to me, that is what becomes more valuable about working with a really good human advisor. First of all, one thing that I think is that it does level the playing field in the sense that you can identify a really good human advisor, not based on list size or platform size. All of those can be improved through the use of AI, but really on good instinct, just that general sense of safety and comfort you have with them.
Is this somebody that I can share? It’s that relationship piece. I know there are people out there, and AI is their best friend. I am not. I’m like, I still know it’s a robot, like you said. I’m not building a strong relationship with my AI. I build a very strong relationship with the humans that I do business with. Would you say there’s anything else that’s going to become more valuable, even as AI improves?
Relationships are going to be number one because we are a species that requires connection. I do not think that’s ever going to change. Maybe that’s an evolutionary thing that’s going to take a long time. Our survival has been based on connection. Connection is going to be critical to us. That’s the piece that AI cannot do well and should not be expected to do at all.
That brings us to the end of this episode, which I went around a few corners I did not think we were going to go to. I think we definitely honored our promise of fully exploring why somebody might pay for an expert when AI can give them an answer. I’m going to say if this episode has made you question any of your assumptions around AI, around what human advisors can and cannot do for you, you might be really interested in Gwen’s one really good question email because it’s an invitation to reflect on a weekly basis on some of those things that you think you know but maybe do not know for a fact whether you know.
It is the simplest process ever. Gwen sends one question each week to help you step back and think differently. You choose whether or not to respond. The emails come every Friday until you decide to unsubscribe. If you’d like to sign up, go to www.tonyakubo.com/Question, and it’s an easy process. We’ll see you next time.
Mentioned in This Episode
About Your Hosts
Gwen Bortner has spent four decades advising executives and entrepreneurs in 45+ industries. She helps women succeed in business without sacrificing happiness by identifying their true desires and aligning their business functions. She spots overlooked bottlenecks and crafts efficient plans toward sustainable success that center your values and priorities. Known for her unique approach to problem-solving and accountability through the G.E.A.R.S. framework, Gwen empowers clients to achieve their definition of success without sacrificing what matters most.
Tonya Kubo is a marketing strategist and community builder who helps entrepreneurs build thriving online communities. As co-host of The Business You Really Want and Chief Marketing and Operations Officer (CMOO) at Everyday Effectiveness, she keeps conversations on track and ensures complex business concepts are accessible to everyone. A master facilitator with 18+ years of experience in online community building, Tonya takes a people-first approach to marketing and centers the human experience in all she does.
0 Comments